Laminar-specific and developmental expression of aquaporin-4 in the mouse hippocampus

Mice deficient in the water channel aquaporin-4 (AQP4) demonstrate increased seizure duration in response to hippocampal stimulation as well as impaired extracellular K+ clearance. However, the expression of AQP4 in the hippocampus is not well described. In this study, we investigated (i) the developmental, laminar and cell-type specificity of AQP4 expression in the hippocampus; (ii) the effect of Kir4.1 deletion on AQP4 expression; and (iii) performed Western blot and RT-PCR analyses. AQP4 immunohistochemistry on coronal sections from wild-type (WT) or Kir4.1-/- mice revealed a developmentally-regulated and laminar-specific pattern, with highest expression in the CA1 stratum lacunosum-moleculare (SLM) and the molecular layer (ML) of the dentate gyrus (DG). AQP4 was colocalized with the glial markers glial fibrillary acidic protein (GFAP) and S100β in the hippocampus, and was also ubiquitously expressed on astrocytic endfeet around blood vessels. No difference in AQP4 immunoreactivity was observed in Kir4.1-/- mice. Electrophysiological and postrecording RT-PCR analyses of individual cells revealed that AQP4 and Kir4.1 were co-expressed in nearly all CA1 astrocytes. In NG2 cells, AQP4 was also expressed at the transcript level. This study is the first to examine subregional AQP4 expression during development of the hippocampus. The strikingly high expression of AQP4 in the CA1 SLM and DG ML identifies these regions as potential sites of astrocytic K+ and H2O regulation. These results begin to delineate the functional capabilities of hippocampal subregions and cell types for K+ and H2O homeostasis, which is critical to excitability and serves as a potential target for modulation in diverse diseases.

Analysis of phosphorylation-dependent modulation of Kv1.1 potassium channels

The voltage-gated potassium channel Kv1.1 contains phosphorylation sites for protein kinase A (PKA) and protein kinase C (PKC). To study Kv1.1 protein expression and cellular distribution in regard to its level of phosphorylation, the effects of PKA and PKC activation on Kv1.1 were investigated in HEK 293 cells stably transfected with Kv1.1 (HEK 293/1). Without kinase activation, HEK 293/1 cells carry unphosphorylated Kv1.1 protein in the plasma membranes, whereas large amounts of phosphorylated and unphosphorylated Kv1.1 protein were located intracellularly. Activation of PKA resulted in phosphorylation of intracellular Kv1.1 protein, followed by a rapid translocation of Kv1.1 into the plasma membrane. Patch-clamp analysis revealed an increase in current amplitude upon PKA activation and demonstrated differences in the voltage dependence of current activation between unphosphorylated and phosphorylated Kv1.1 channels. In contrast to PKA, even prolonged activation of PKC did not lead to direct phosphorylation of Kv1.1, but induced Kv1.1 protein synthesis. Thus, protein kinases have direct and indirect effects on the functional expression of voltage-gated potassium channels. Our data suggest that the synergistic action of protein kinases may play an important role in the fine-tuning of Kv channel function.

Modulation of voltage-gated K(+) channels Kv11 and Kv1 4 by forskolin

Forskolin (FSK) affects voltage-gated K(+) (Kv) currents in different cell types, but it is not known which of the various subunits form FSK-sensitive Kv channels. We compared the effect of the compound at Kv1.1 and Kv1.4 channels ectopically expressed in HEK 293 cells. Low FSK concentrations induced a phosphorylation-dependent potentiation of Kv1.1 currents. At higher concentrations, this effect was superimposed by a fast, cAMP-independent channel block. Kv1.4 currents were inhibited with lower potency by FSK but were not modified by phosphorylation. The variable effect of the compound might help to distinguish between Kv subunits expressed by native cells.

Heterogeneity in the Membrane Current Pattern of Identified Glial Cells in the Hippocampal Slice

Glial cells, acutely isolated or in tissue culture, have previously been shown to express a variety of voltage-gated channels. To resolve the question whether such channels are also expressed by glial cells in their normal cellular environment, we have applied the patch-clamp technique to study glial cells in hippocampal slices of 10 - 12-day-old mice. Based on the membrane current pattern, we distinguished four glial cell types. One was characterized by passive, symmetrical K+ currents activated in depolarizing and hyperpolarizing directions. A second population showed a similar current pattern, but with a marked decay of the current during the 50-ms voltage jumps. In a third population, the decaying passive currents were superimposed with a delayed rectifier outward current and, in some cases, with a slow inward current activated by depolarization. The fourth population expressed delayed rectifying outward currents, an inward rectifier K+ current and fast inward currents activated by depolarization. To unequivocally identify the glial cells we combined electrophysiological and ultrastructural characterizations. Therefore, cells were filled with the fluorescent dye lucifer yellow during characterization of their membrane currents, the fluorescence of the dye was used to convert diaminobenzidine to an electron-dense material, and subsequently slices were inspected in the electron microscope. Recordings were obtained from cells in the stratum radiatum and were identified as glial by their size, the characteristic chromatin distribution, and the lack of synaptic membrane specializations.

6-Hydroxykynurenic acid and kynurenic acid differently antagonise AMPA and NMDA receptors in hippocampal neurones

6-Hydroxykynurenic acid (6-HKA), a derivative of kynurenic acid (KYNA) extracted from Ginkgo biloba leaves, was tested for its putative glutamate receptor (GluR) antagonism in comparison to the scaffold substance. The patch-clamp method together with fast-application techniques were used to estimate inhibition by 6-HKA and KYNA of agonist binding at NMDA and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors (NMDARs and AMPARs) of CA1 pyramidal neurones. 6-Hydroxykynurenic acid proved to be a low-affinity antagonist. When comparing with KYNA, 6-HKA was less potent at NMDARs (IC(50) = 136 versus 59 microM), but showed a higher affinity to AMPARs (K(B) = 22 versus 172 microM). The replacement of 6-HKA and KYNA by glutamate was investigated on outside-out patches. Both antagonists competitively inhibited AMPAR responses and displayed fast unbinding kinetics, but the derivative was significantly slower displaced than KYNA (tau = 1.63 versus 1.22 ms). Our findings demonstrate that 6-hydroxylation considerably changes the pharmacological profile of KYNA. Among the 6-derivatives of KYNA, 6-HKA shows the highest affinity to AMPARS: Despite its relatively low lipophily, these properties might be of clinical relevance under conditions that compromise the integrity of the blood-brain barrier. Furthermore, 6-HKA should be a useful tool to analyse glutamate-mediated synaptic responses.

Distribution of P2X receptors on astrocytes in juvenile rat hippocampus

Recent evidence suggested that ATP acting via ionotropic (P2X) and metabotropic (P2Y) purinergic receptors might be involved in signaling between glial cells and within glial-neuronal networks. In contrast to their neuronal counterpart, the identity of P2X receptors in CNS glial cells is largely unknown. In the present study, antibodies recognizing the subunits P2X1-P2X7 were applied together with the astroglial marker S100beta and nuclear labeling with Hoechst 33342 to investigate semiquantitatively the distribution of the whole set of P2X receptors in astrocytes of the juvenile rat hippocampus. Expression of P2X1-P2X4, P2X6, and P2X7 subunits was observed in astrocytes of various hippocampal subregions, but the cells were completely devoid of P2X5 protein. S100beta-positive cells expressing subunits P2X3-P2X7 occurred evenly in the different subfields, while P2X1- and P2X2-positive astrocytes were distributed more heterogeneously. The staining pattern of P2X subunits also differed at the subcellular level. Antibodies against P2X2 and P2X4 labeled both astroglial cell bodies and processes. Immunoreactivity for P2X1 and P2X6 was mainly confined to somatic areas of S100beta-positive cells, whereas the subunit P2X3 was primarily localized along astroglial processes. Knowledge of the distribution of P2X receptors might provide a basis for a better understanding of their specific role in cell-cell signaling.

Characterization of the anticonvulsant and neuroprotectant BIIR 561 CL in vitro: effects on native and recombinant alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors

BIIR 561 CL is a novel blocker of AMPA receptors and voltage-dependent sodium channels. In this study we further describe the effects of BIIR 561 CL on AMPA receptor-mediated membrane currents in rodent neurons, as well as in cells expressing recombinant human GluR1/2 receptors in more detail. BIIR 561 CL suppressed responses to kainate in neuronal cultures from rat cortex with an IC50 of 9.8 microM. Similar effects were observed using acutely dissociated neurons from the CA1 region of rat hippocampus (IC50 = 9.5 microM). Inhibition of kainate responses by BIIR 561 CL was prevented by preapplication of GYKI 53655, suggesting that both non-competitive inhibitors bind to a common site of the receptor. The effect of 10 microM BIIR 561 CL on kainate-induced currents was dependent on extracellular pH, with more pronounced block (84.1%) under acidic conditions (pHextern=6.4), compared to only 30.1% at a pHextern of 8.4. Thus, it can be hypothesized that BIIR 561 CL inhibits AMPA receptors in ischaemic brain regions more effectively than in healthy tissue. BIIR 561 CL inhibited responses to 1 mM glutamate in cells expressing recombinant human GluR1/2 receptors with similar potency, as compared to kainate responses in rat neurons (IC50=17.3 microM). The reference compound NBQX had an IC50 of 25.2 nM. None of the two compounds affected the glutamate-induced receptor desensitization at any tested concentration. The block by BIIR 561 CL was not use-dependent and had fast on- and off-kinetics (tauon=6.8 s; tauoff=1.3 s in hGluR1/2 receptors with 30 microM BIIR 561 CL). Thus, BIIR 561 CL can be anticipated to have a promising profile for the treatment of neurological disorders like brain ischaemia and head trauma.

Functional and molecular properties of human astrocytes in acute hippocampal slices obtained from patients with temporal lobe epilepsy

PURPOSE

The specific role of glial cells in epilepsy is still elusive. In this study, functional properties of astrocytes were investigated in acute hippocampal brain slices obtained from surgical specimens of patients with drug-resistant temporal lobe epilepsy (TLE).

METHODS

The patch-clamp technique together with a single-cell reverse transcription-polymerase chain reaction approach were used to combine functional and molecular analysis in the same individual cell in situ.

RESULTS

In patients with Ammon's horn sclerosis, the glial current patterns resembled properties of immature astrocytes in rodent hippocampus. Depolarizing voltage steps activated delayed rectifier and transient K+ currents as well as tetrodotoxin-sensitive Na+ currents. Hyperpolarizing voltages elicited inward rectifier K+ currents. Comparative recordings were made in astrocytes from patients with lesion-associated TLE that lacked significant histopathological hippocampal alterations. The inward rectifier K+ current density was significantly smaller in astrocytes from the sclerotic group compared with lesion-associated TLE patients.

CONCLUSIONS

During normal development of rodent brain, astroglial inward rectification gradually increases. It thus appears that astrocytes in human sclerotic tissue reexpress an immature current pattern. Reduced astroglial inward rectification in conjunction with seizure-induced shrinkage of the extracellular space may lead to impaired spatial K+ buffering. This will result in stronger and prolonged depolarization of glial cells and neurons in response to activity-dependent K+ release and may thus contribute to seizure generation and spread in this particular condition of human TLE.

Modulation of calcium channels by group I and group II metabotropic glutamate receptors in dentate gyrus neurons from patients with temporal lobe epilepsy

PURPOSE

Metabotropic glutamate receptors (mGluRs) might be promising new drug targets for the treatment of epilepsy because the expression of certain mGluRs is regulated in epilepsy and because activation of mGluRs results in distinctive anti- and proconvulsant effects. Therefore, we examined how mGluR activation modulates high-voltage-activated (HVA) Ca2+ channels.

METHODS

Whole-cell patch-clamp recordings were obtained from granule cells and interneuron-like cells acutely isolated from the dentate gyrus of patients with pharmacoresistent temporal lobe epilepsy.

RESULTS

Agonists selective for either group I or group II mGluRs rapidly and reversibly reduced HVA currents in most dentate gyrus cells. These modulatory effects were inhibited by the respective group I and group II mGluR antagonists. The specific Ca2+ channel antagonists nifedipine and omega-conotoxin GVIA potently occluded the effects of group I and II mGluR agonists, respectively, indicating that group I mGluRs acted on L-type channels and group II mGluRs affected N-type channels. About two thirds of the responsive neurons were sensitive either to group I or group II mGluRs, whereas a minority of cells showed effects to agonists of both groups, indicating a variable mGluR expression pattern.

CONCLUSIONS

Group I and group II mGluRs are expressed in human dentate gyrus neurons and modulate L- and N-type HVA channels, respectively. The data shed light on the possible cellular sequelae of the mGluR1 upregulation observed in human epileptic dentate gyrus as well as on possible mGluR-mediated anticonvulsant mechanisms.

Voltage- and gamma-aminobutyric acid-activated membrane currents in the human medulloblastoma cell line MHH-MED-3

The whole-cell patch clamp technique was used to characterize voltage- and neurotransmitter-activated currents in the medulloblastoma cell line MHH-MED-3 and cells from tissue slices and primary cultures of two medulloblastoma biopsies. These preparations revealed similar electrophysiological properties. All tested cells displayed 4-aminopyridine-sensitive delayed rectifying K(+) currents, gamma-aminobutyric acid(A) receptor-mediated Cl(-) currents and most of them inward rectifier K(+) currents. Transient inward currents were mainly carried by low-voltage activated T-type Ca(2+) channels in MHH-MED-3 cells, and tetrodotoxin-sensitive Na(+) channels in cells from the primary culture. From these characteristics we conclude that medulloblastoma cells share physiological features with developing cerebellar granule cells at an immature stage.

Astrocytes in the hippocampus of patients with temporal lobe epilepsy display changes in potassium conductances

Functional properties of astrocytes were investigated with the patch-clamp technique in acute hippocampal brain slices obtained from surgical specimens of patients suffering from pharmaco-resistant temporal lobe epilepsy (TLE). In patients with significant neuronal cell loss, i.e. Ammon's horn sclerosis, the glial current patterns resembled properties characteristic of immature astrocytes in the murine or rat hippocampus. Depolarizing voltage steps activated delayed rectifier and transient K+ currents as well as tetrodotoxin-sensitive Na+ currents in all astrocytes analysed in the sclerotic human tissue. Hyperpolarizing voltages elicited inward rectifier currents that inactivated at membrane potentials negative to -130 mV. Comparative recordings were performed in astrocytes from patients with lesion-associated TLE that lacked significant histopathological hippocampal alterations. These cells displayed stronger inward rectification. To obtain a quantitative measure, current densities were calculated and the ratio of inward to outward K+ conductances was determined. Both values were significantly smaller in astrocytes from the sclerotic group compared with lesion-associated TLE. During normal development of rodent brain, astroglial inward rectification gradually increases. It thus appears reasonable to suggest that astrocytes in human sclerotic tissue return to an immature current pattern. Reduced astroglial inward rectification in conjunction with seizure-induced shrinkage of the extracellular space may lead to impaired spatial K+ buffering. This will result in stronger and prolonged depolarization of glial cells and neurons in response to activity-dependent K+ release, and may thus contribute to seizure generation in this particular condition of human TLE.

Developmental regulation of AMPA-receptor properties in CA1 pyramidal neurons of rat hippocampus

AMPA-receptor (AMPA-R) currents were recorded from CA1 pyramidal neurons in situ and after acute isolation from the hippocampus of 3- to 45-day-old rats. Membrane currents were analyzed by combining the patch clamp method with fast application techniques. The complete block of receptor currents by GYKI 53655 and the absence of modulation by Concanavalin A indicated that the cells exclusively expressed non-NMDA glutamate receptors of the AMPA subtype while functional kainate receptors could not be detected. The lowest sensitivity to kainate and NBQX was observed at postnatal day (p) 18. These changes might reflect a lower abundance of GluR1 at that developmental stage. A decrease of potentiation of receptor currents by cyclothiazide (CTZ), an acceleration of the recovery from CTZ potentiation and a faster and more complete desensitization of glutamate-evoked currents suggest an up-regulation of flop splice variants with increasing age. These functional data indicate that AMPA-R expression in CA1 pyramidal neurons varies during postnatal development which can be expected to influence the kinetics of synaptic transmission and the excitotoxic vulnerability as well.

Ion channels in glial cells

Functional and molecular analysis of glial voltage- and ligand-gated ion channels underwent tremendous boost over the last 15 years. The traditional image of the glial cell as a passive, structural element of the nervous system was transformed into the concept of a plastic cell, capable of expressing a large variety of ion channels and neurotransmitter receptors. These molecules might enable glial cells to sense neuronal activity and to integrate it within glial networks, e.g., by means of spreading calcium waves. In this review we shall give a comprehensive summary of the main functional properties of ion channels and ionotropic receptors expressed by macroglial cells, i.e., by astrocytes, oligodendrocytes and Schwann cells. In particular we will discuss in detail glial sodium, potassium and anion channels, as well as glutamate, GABA and ATP activated ionotropic receptors. A majority of available data was obtained from primary cell culture, these results have been compared with corresponding studies that used acute tissue slices or freshly isolated cells. In view of these data, an active glial participation in information processing seems increasingly likely and a physiological role for some of the glial channels and receptors is gradually emerging.

Developmental changes in the expression of Shaker- and Shab-related K(+) channels in neurons of the rat trigeminal ganglion

We have investigated properties of voltage-gated K(+) channels in neurons of the pre- and postnatal rat trigeminal ganglion (TG). To correlate functional data with information on gene expression of Shaker- and Shab-related channels in these pseudo-unipolar neurons, the patch-clamp technique was combined with the single-cell reverse transcription-polymerase chain reaction (RT-PCR). A majority (80%) of prenatal TG neurons possessed only sustained delayed rectifier currents with half-maximal current inactivation at -30 mV. In the postnatal cells, steady-state inactivation of sustained currents occurred at more negative voltages (half-maximal inactivation at -58 mV). About 65% of the postnatal cells displayed a transient outward component in addition to the sustained currents. With increasing age, the sensitivity of sustained currents to 4-aminopyridine (4-AP) decreased significantly. The Shaker channel toxins, alpha-dendrotoxin and agitoxin-2 (50 and 10 nM), were much less effective. Discrimination between both stages with tetraethylammonium chloride (5 mM) was not possible since the currents were reduced generally by about 50%. After recording, the cell content was harvested and single-cell RT-PCR was performed to compare K(+) current properties and mRNA expression within the same cell. Most cells simultaneously expressed several different Shaker- and Shab-like transcripts. At postnatal day 14, the frequency of cells carrying transcripts encoding Kv1.1 decreased. Detailed analysis revealed a higher 4-AP sensitivity of TG neurons expressing Kv1.1 transcripts.

Lesion-induced changes of electrophysiological properties in astrocytes of the rat dentate gyrus

Reorganization of the adult dentate gyrus following unilateral entorhinal cortex lesion (ECL) is a well-established model for studying mechanisms of trauma-induced neuronal plasticity. The lesion induces deafferentiation of the outer molecular layer, which is accompanied by a strong astroglial reaction. This glial response is thought to contribute to subsequent repair processes, but the underlying mechanisms are poorly understood. In this study we addressed the question whether denervation leads to modifications in the electrophysiological properties of astrocytes, assuming that such changes might be involved in the remodeling of neural circuitry. Patch-clamp recordings were obtained from astrocytes in the dentate gyrus of adult rats that underwent ECL and compared to corresponding data from control animals. We observed a significant reduction of inward rectifier K(+) current densities, a positive shift of resting potentials, and an increase in input resistance in astrocytes of the denervated molecular layer. Current densities were reduced between 6 and 19 days postlesion (dpl), reaching a minimum at 10 dpl. Voltage-gated outward K(+) currents were not affected by the lesion. Inward rectifier K(+) currents increase with maturation in astrocytes. Thus, our results provide evidence that, following ECL, mature astrocytes dedifferentiated and readapted an immature current pattern. Presumably, these changes lead to stronger and prolonged depolarization of glial cells and neurons in response to activity-dependent K(+) release, which in turn might enhance the synthesis of neurotrophic factors and contribute to a permissive environment for neuronal reorganization.

Two electrophysiologically distinct types of granule cells in epileptic human hippocampus

We investigated the electrophysiology of morphologically identified human granule cells with conventional current-clamp recordings. Slices were prepared from 14 human epileptic sclerotic hippocampi. Granule cells appeared to have a diverse electrophysiology. Each cell was distinguished by the shape of the afterhyperpolarization following single action potentials. Two types could be discerned: type I afterhyperpolarizations were monophasic and brief (typically 10-40 ms), whilst type II afterhyperpolarizations were biphasic and long (typically 50-100 ms). The two types also differed in their repetitive firing behaviour and action potential morphology: type I cells had significantly weaker spike frequency adaptation, lower action potential amplitude and smaller action potential upstroke/downstroke ratio. Thus, the firing pattern of type I cells resembled that of rodent dentate interneurons. In contrast, the corresponding parameters of type II cells were comparable to rodent dentate granule cells. Despite the distinct firing patterns, membrane properties were not different. The two types of cells also differed in their synaptic responses to stimulation of the perforant path. At strong suprathreshold stimulation intensity, type I cells always generated multiple action potentials, whereas type II cells usually spiked once only. Slow inhibitory postsynaptic potentials were not detected in type I neurons, but were easily identified in type II neurons. Extracellular recordings of perforant path-evoked field potentials in the cell layer confirmed that the majority of granule cells showed multiple discharges even when we recorded simultaneously from a type II cell that generated one action potential only. The morphology of both types of cells was characteristic of what has been described for primate dentate granule cells. Based on comparisons with previous studies on rodent and human granule cells, we tentatively hypothesize that: (i) the majority of granule cells from sclerotic hippocampus display an hyperexcitable epileptogenic electrophysiology; (ii) there is a subset of granule cells whose electrophysiology is preserved and is more comparable to granule cells from non-epileptic hippocampus.

Late onset and increasing expression of the gap junction protein connexin30 in adult murine brain and long-term cultured astrocytes

In rat brain, expression of the gap junction protein connexin30 increased during the first 3 weeks after birth and reached its maximum after 4 weeks, as shown by analysis with specific connexin30 antibodies. This contrasts with the prenatal onset of connexin43 expression. On cryosections of rat brain, connexin30 immunoreactivity was found near blood vessels and in ependymal as well as in leptomeningeal cells. Expression in the neuropil was first noticed 3 weeks after birth, showing the same spatial pattern of immunoreactivity as connexin43. This late onset of connexin30 expression in astrocytes was also seen in long-term glial cell cultures, where connexin30 was coexpressed with the astrocytic marker proteins S-100beta and glial fibrillary acid protein. In acute brain slices, connexin30 immunofluorescent signals were detected on processes of functionally identified astrocytes. Thus, our results show that connexin30 is expressed in three different cell types of the rodent brain. The late onset of connexin30 expression in astrocytes suggests that this gap junctional protein fulfills a role in intercellular communication among mature astrocytes.

Effects of phenytoin, carbamazepine, and gabapentin on calcium channels in hippocampal granule cells from patients with temporal lobe epilepsy

PURPOSE

The anticonvulsants phenytoin (PHT), carbamazepine (CBZ), and gabapentin (GBP) are commonly used in the treatment of temporal lobe epilepsy. Ca2+ current modulation has been proposed to contribute to the antiepileptic activity of these drugs. The purpose of this study was to determine the effects of these anticonvulsants on voltage-dependent calcium channels in pathologically altered neurons from patients with chronic temporal lobe epilepsy.

METHODS

Acutely isolated human hippocampal granule cells were examined by using the whole-cell configuration of the patch-clamp technique.

RESULTS

PHT and CBZ produced a reversible, concentration-dependent inhibition of high-voltage-activated (HVA) Ca2+ currents without affecting voltage-dependent activation. The concentration-response curves of PHT and CBZ indicated maximal inhibition of 35 and 65%, respectively, with half-maximal inhibition being obtained at 89 and 244 microM, respectively. At therapeutic cerebrospinal fluid (CSF) concentrations, HVA currents were not significantly altered by PHT and CBZ. However, PHT but not CBZ showed a reduction of HVA currents of 16% at a therapeutic whole-brain concentration of 80 microM. In contrast to CBZ, PHT produced a small hyperpolarizing shift in the voltage dependence of steady-state inactivation. PHT, 80 microM, shifted the potential of half-maximal inactivation by -3.1 +/- 0.5 mV (p < 0.05). GBP, which was recently found to bind to the alpha2delta subunit of a neuronal Ca2+ channel, showed no modulation of Ca2+ conductances.

CONCLUSIONS

These results suggest that, in contrast to GBP and CBZ, modulation of postsynaptic Ca2+ channels can contribute to the anticonvulsant action of PHT in human hippocampal granule cells.

Analysis of AMPA receptor properties during postnatal development of mouse hippocampal astrocytes

Glial cells in the mammalian brain express various types of voltage- and ligand-gated ion channels, including glutamate receptors (GluRs) of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-subtype. In the present study we followed developmental changes in the functional properties of AMPA receptor (AMPA-R) channels expressed by astrocytes of the mouse hippocampus between postnatal days (P) 5-35 to learn more about the physiological significance of these glial receptors. A fast concentration clamp technique was applied to cells acutely isolated from the CA1 stratum radiatum subregion to quantitatively analyze rapidly activating and desensitizing receptor responses. The equilibrium responses of glutamate and kainate differed between P5 and P12. Although the maximum current induced by kainate was almost the same at all developmental stages, a steep rise in the maximum glutamate response was observed within the same time range. Between P5 and P12 there was an increase in the potentiation of AMPA-R currents with cyclothiazide (CTZ); at the same time, the dissociation kinetics of CTZ became significantly slower. These changes in the pharmacological properties suggested a variation in splice variant expression. With proceeding maturation, we observed an increase in the degree of desensitization of the glutamate- and AMPA-induced receptor currents. In addition to the shift in flip/flop splicing, these findings could hint at a developmental regulation of RNA editing in the arginine/glycine site. Altogether, the present results demonstrate changes in astrocytic AMPA-R functioning early in postnatal development, although after P12 the receptor properties remained almost constant. Although the overall Ca2+ permeability did not vary during development, the prolonged receptor opening in the early postnatal period causes an enhanced Na+/Ca2+ influx into the immature astrocytes. This could influence glial proliferation and differentiation during CNS ontogenesis.

AMPA receptor subunits expressed by single astrocytes in the juvenile mouse hippocampus

The subunit composition of native AMPA receptor (AMPA-R) channels was recently described in several neuronal cell types but less information is available on glial cells. Evidence from recombinant receptor studies suggests that the expression of distinct subunits determines the specific functional properties of the receptor channel. In the present study, we combined the patch clamp technique with the reverse transcription-polymerase chain reaction (RT-PCR) to correlate the expression of gene transcripts with functional properties of AMPA-R in single identified glial cells of the hippocampus. The cells were freshly isolated from the stratum radiatum of the CA1 subregion. We focused on cells expressing AMPA-R with an intermediate Ca2+ permeability which were identified as immature astrocytes due to their morphological, immunocytochemical and electrophysiological characteristics. After recording, the cells were harvested and RT-PCR was performed with the same individual cell to investigate the composition of their AMPA-R transcripts. Our results suggest the expression of a heteromeric subunit architecture. In all cells, the GluR2 subunit was present, which is known to confer a low Ca2+ permeability to the receptor complex. Most frequently, we met co-expression of GluR2 and GluR4. This study demonstrates that astrocytes in the hippocampus express a distinct AMPA-R subunit composition which differs from neurons. The glial receptors might be involved in the modulation of gene expression as well as the regulation of proliferation and differentiation.

News on glutamate receptors in glial cells

Glutamate (Glu) receptors convey most of the excitatory synaptic transmission in the mammalian CNS. Distinct Glu-receptor genes and different subtypes of glutamate-activated channels are expressed ubiquitously throughout the developing and mature brain in the two major macroglial cell types, astrocytes and oligodendrocytes. These glial receptors are found in acutely isolated cells and in brain slices, and are therefore functional in vivo. Glutamate receptors in glial cells are activated during neuronal activity, and their activation modulates gene expression in astrocytes and oligodendrocytes. The proliferation and differentiation of glial precursor cells are also regulated by activation of Glu receptors, suggesting that the excitatory transmitter might be one of the environmental signals that regulate glial-cell development.

Identified glial cells in the early postnatal mouse hippocampus display different types of Ca2+ currents

Based on their typical pattern of membrane currents, four populations of glial cells could be identified in thin brain slices of the postnatal hippocampus. In the present study, we applied the patch-clamp technique to glial cells in the hippocampal CA1 region, which are characterized by a complex pattern of different Na+ and K+ currents ("complex" cells). These cells were identified as non-neuronal cells, most likely astrocytes, by their glutamine synthetase immunoreactivity. Two types of glial Ca2+ currents could be identified that differed in their kinetics and pharmacological properties. A low-voltage activated (LVA), fast inactivating component was activated at membrane potentials positive to -60 mV and reached maximum current amplitudes at about -20 mV. This current was sensitive to amiloride and thus displayed properties of neuronal LVA currents. The threshold potential of the second Ca2+ current component was at about -40 mV, and peak currents were observed at 0 mV. In contrast to the LVA component, the inactivation of these high-voltage activated (HVA) currents slowed down with increasing depolarizations. This current was sensitive to low concentrations of Cd2+ but was not affected by amiloride. A small fraction of the HVA currents was sensitive to nifedipine, and omega-conotoxin GVIA (omega-CgTx) was also found to reduce the glial HVA component. The study provides electrophysiological and pharmacological characterization of different types of Ca2+ currents in gray matter glial cells in situ.

Developmental regulation of Na+ and K+ conductances in glial cells of mouse hippocampal brain slices

The relative contribution of voltage activated Na+ and K+ currents to the whole cell current pattern of hippocampal glial cells was analyzed and compared during different stages of postnatal maturation. The patch-clamp technique was applied to identified cells in thin brain slices obtained from animals between postnatal day 5 and 35 (p5-35). We focused on a subpopulation of glial cells in the CA1 stratum radiatum which most probably represents a pool of immature astrocytes, termed "complex" cells. These cells could not be labelled by O1/O4 antibodies, but some of the older cells were positively stained for glial fibrillary acidic protein (GFAP). In the early postnatal days, the current pattern of the "complex" cells was dominated by two types of K+ outward currents: a delayed rectifier and a transient component. In addition, all cells expressed significant tetrodotoxin (TTX)-sensitive Na+ currents. During maturation, the contribution of delayed rectifier and A-type currents significantly decreased. Furthermore, almost all cells after p20 lacked Na+ currents. This down-regulation of voltage gated Na+ and K+ outward currents was accompanied by a substantial increase in passive and inward rectifier K+ conductances. We found increasing evidence of electrical coupling between the "complex" cells with continued development. It is concluded that these developmental changes in the electrophysiological properties of "complex" glial cells could be jointly responsible for the well known impaired K+ homeostasis in the early postnatal hippocampus.

Glial cells in the mouse hippocampus express AMPA receptors with an intermediate Ca2+ permeability

Recently, we could demonstrate the 'complex' glial cells in mouse hippocampal slices express glutamate receptor changes of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)/kainate subtypes. In the present study, we further characterized this glial receptor. Since voltage-clamp control is imperfect and diffusion barriers hinders the quantitative analysis of the receptor currents in situ, the patch-clamp technique was applied to glial cells acutely isolated from the mouse hippocampal CA1 stratum radiatum subregion. A concentration-clamp technique was used which enabled very fast exchange of the extracellular solutions. Thus, it was possible to characterize the transient receptor currents with high time resolution. Application of L-glutamate, AMPA and L-homocysteate induced rapidly activating and fast desensitizing receptor currents in the suspended glial cells. In contrast, kainate induced non-desensitizing currents. The corresponding dose-response curve revealed a half-maximum of current activation at 350 microM. The current/voltage relationship of the kainate-evoked response was linear with a reversal potential at approximately 9 mV. Analysis of the reversal potential in solutions containing high concentrations of CaCl2 confirmed earlier in situ data by demonstrating significant Ca2+ permeability of the glial glutamate receptor channels in the hippocampus. The kainate-induced receptor currents were markedly increased by cyclothiazide, a substance which selectively potentiates glutamate receptors of the AMPA subtype. We conclude that glial cells of the juvenile hippocampus mainly express heteromeric high-affinity AMPA receptors. Most probably, the receptor channels are assembled from the low Ca(2+)-permeable glutamate receptor-2 subunit together with Ca(2+)-permeable AMPA-preferring subunits.

Properties of voltage-activated Na+ and K+ currents in mouse hippocampal glial cells in situ and after acute isolation from tissue slices

In the present study, we were interested in a quantitative analysis of voltage-activated channels in a subpopulation of hippocampal glial cells, termed "complex" cells. The patch-clamp technique in the whole-cell mode was applied to identified cells in situ and to glial cells acutely isolated from tissue slices. The outward current was composed of two components: a sustained and a transient current. The transient K+ channel had electrophysiological and pharmacological properties resembling those of the channel through which the A-currents pass. In addition, this glial A-type current possessed a significant Ca2+ dependence. The current parameters determined in situ or in isolated cells corresponded well. Due to space clamp problems in situ, properties of voltage-dependent Na+ currents were only analysed in suspended glial cells. The tetrodotoxin (TTX) sensitivity and the stationary and kinetic characteristics of this current were similar to corresponding properties of hippocampal neurons. These quantitative data demonstrate that at an early postnatal stage of central nervous system maturation, glial cells in situ express a complex pattern of voltage-gated ion channels. The results are compared to findings in other preparations and the possible consequences of transmitter-mediated channel modulation in glial cells are discussed.

Properties of GABA and glutamate responses in identified glial cells of the mouse hippocampal slice

In this study, the patch-clamp technique was applied to brain slices to test for the presence of GABAA and glutamate receptors in glial cells of an intact tissue preparation, the hippocampus from 9-12 day old mice. Two types of glial cells were studied in the CA1 stratum pyramidale, termed passive and complex cells, which were distinct by their characteristic pattern of voltage-dependent currents. Both cell types were previously identified as glial by combining electrophysiology with ultrastructural inspection (Steinhüser et al., 1992, Eur J Neurosci 4:472-484). A subpopulation of passive cells was positive, all complex cells were negative for immunocytochemical staining against glial fibrillary acidic protein, a marker of mature astrocytes. In both cell types, GABA activated currents compatible with GABAA-receptor mediated responses. The glutamate response in complex and in most of the passive cells was mediated by a ligand-gated ion channel and closely matched the pharmacology of the kainate receptor. Activation of glutamate receptors led to a transient decrease of the resting K+ conductance in complex cells and to an irreversible decrease in the passive cells. In three passive cells, glutamate-activated currents were most likely dominated by an electrogenic uptake. In a small group of passive cells NMDA-activated currents were observed. This study provides evidence that glial cells from an intact tissue express receptors for the most abundant transmitters in the central nervous system, glutamate, and GABA.

Kainate activates Ca(2+)-permeable glutamate receptors and blocks voltage-gated K+ currents in glial cells of mouse hippocampal slices

Glial cells in the CA1 stratum radiatum of the hippocampus of 9- to 12-day-old mice show intrinsic responses to glutamate due to the activation of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA)/kainate receptors. In the present study we have focused on a subpopulation of the hippocampal glial cells, the "complex" cells, characterized by voltage-gated Na+ and K+ channels. Activation of glutamate receptors in these cells led to two types of responses, the activation of a cationic conductance, and a longer-lasting blockade of voltage-gated K+ channels. In particular, the transient (inactivating) component of the outwardly rectifying K+ current was diminished by kainate. Concomitantly, as described in Bergmann glial cells, kainate also elevated cytosolic Ca2+. This increase was due to an influx via the glutamate receptor itself. In contrast to Bergmann glial cells, the cytosolic Ca2+ increase was not a link to the K+ channel blockade, since the blockade occurred in the absence of the Ca2+ signal and, vice versa, an increase in cytosolic Ca2+ induced by ionomycin did not block the transient K+ current. We conclude that glutamate receptor activation leads to complex and variable changes in different types of glial cells; the functional importance of these changes is as yet unresolved.

Proteolytic enzymes do not destroy the N-methyl-D-aspartate (NMDA) sensitivity of acutely isolated hippocampal CA1 and CA3 neurons from postnatal rats

Using whole cell recordings in combination with the concentration clamp technique it was shown that even enzymatically isolated hippocampal CA1 and CA3 neurons exhibit N-methyl-D-aspartate (NMDA) and L-aspartate currents. These currents were completely blocked by 0.5 mM Mg2+ or partially blocked (72.7 +/- 2.4%) by DL-2-amino-5-phosphonovalerate (50 microM) whereas 10 microM glycine (Gly) potentiated the NMDA responses (3.1 +/- 1.6-fold). A half-maximum dose of 55 microM was calculated from the sigmoid NMDA dose-response curve in the presence of 10 microM Gly. The amplitudes of these currents did not depend on the type of the proteolytic enzyme used for cell isolation.

Properties of the fast sodium channels in pyramidal neurones isolated from the CA1 and CA3 areas of the hippocampus of postnatal rats

The fast sodium (Na)current of hippocampal neurones was recorded using the intracellular perfusion method. Neurones were isolated from the CA1 and CA3 hippocampal subregions separately, after treatment of the tissue with trypsin. There were no differences between the current-voltage (I-V) characteristics of CA1 and CA3 neurones. In contrast to this, the steady-state inactivation (h infinity) of both types of neurones was significantly different. Additionally, there were two subpopulations of CA1 neurones, which showed different courses of h infinity. Compared with CA1 neurones, the steady-state activation and inactivation curves of CA3 neurones overlapped much more in the potential range -80 mV to -50 mV. These results are consistent with the well-known fact that CA3 neurones show spontaneous burst activity, while CA1 cells do not. The time constant of activation (tau m) depended upon the membrane potential in the same way for all CA3 neurones investigated. However, there were two different subpopulations of CA3 cells, showing different voltage dependence of the time constant of inactivation. We conclude that these differences reflect two types of pyramidal cells within the same subregion.