Electrophysiological properties of ependymal cells (radial glia) in dorsal cortex of the turtle, Pseudemys scripta

A Lifetime's Adventure in Extracellular K+ Regulation: The Scottish Connection

In a career that has spanned 45 years and shows no signs of slowing down, Dr Bruce Ransom has devoted considerable time and energy to studying regulation of interstitial K⁺. When Bruce commenced his studies in 1969 virtually nothing was known of the functions of glial cells, but Bruce’s research contributed to the physiological assignation of function to mammalian astrocytes, namely interstitial K⁺ buffering. The experiments that I describe in this review concern the response of the membrane potential (Em) of in vivo cat cortical astrocytes to changes in [K⁺]_o, an experimental manoeuvre that was achieved in two different ways. The first involved recording the Em of an astrocyte while the initial aCSF was switched to one with different K⁺, whereas in the second series of experiments the cortex was stimulated and the response of the astrocyte Em to the K⁺ released from neighbouring neurons was recorded. The astrocytes responded in a qualitatively predictable manner, but quantitatively the changes were not as predicted by the Nernst equation. Elevations in interstitial K⁺ are not sustained and K⁺ returns to baseline rapidly due to the buffering capacity of astrocytes, a phenomenon studied by Bruce, and his son Chris, published 27 years after Bruce’s initial publications. Thus, a lifetime spent investigating K⁺ buffering has seen enormous advances in glial research, from the time cells were identified as ‘presumed’ glial cells or ‘silent cells’, to the present day, where glial cells are recognised as contributing to every important physiological brain function.

Radial glia in the proliferative ventricular zone of the embryonic and adult turtle, Trachemys scripta elegans

To better understand the role of radial glial (RG) cells in the evolution of the mammalian cerebral cortex, we investigated the role of RG cells in the dorsal cortex and dorsal ventricular ridge of the turtle, Trachemys scripta elegans. Unlike mammals, the glial architecture of adult reptile consists mainly of ependymoradial glia, which share features with mammalian RG cells, and which may contribute to neurogenesis that continues throughout the lifespan of the turtle. To evaluate the morphology and proliferative capacity of ependymoradial glia (here referred to as RG cells) in the dorsal cortex of embryonic and adult turtle, we adapted the cortical electroporation technique, commonly used in rodents, to the turtle telencephalon. Here, we demonstrate the morphological and functional characteristics of RG cells in the developing turtle dorsal cortex. We show that cell division occurs both at the ventricle and away from the ventricle, that RG cells undergo division at the ventricle during neurogenic stages of development, and that mitotic Tbr2+ precursor cells, a hallmark of the mammalian SVZ, are present in the turtle cortex. In the adult turtle, we show that RG cells encompass a morphologically heterogeneous population, particularly in the subpallium where proliferation is most prevalent. One RG subtype is similar to RG cells in the developing mammalian cortex, while 2 other RG subtypes appear to be distinct from those seen in mammal. We propose that the different subtypes of RG cells in the adult turtle perform distinct functions.

Gap junctions do not underlie changes in whole-cell conductance in anoxic turtle brain

An acute reduction in cell membrane permeability could provide an effective strategy to prolong anoxic survival. A previous study has shown that in the western painted turtle whole-cell neuronal conductance (G(w)) decreases during anoxia, which may be mediated by the activation of adenosine A(1) receptors and calcium. Reduction in G(w) is thought to be the result of ion channel closure, but closure of gap junctions could also be responsible for this phenomenon. In our study, antibody staining of connexin 32 and 43 (Cx32 and Cx43) suggested the presence of gap junctional components in the turtle cortex. To examine if gap junctions were involved in the previously measured anoxic decrease in G(w), neuronal connectivity was assessed through the measurement of whole-cell capacitance (C(w)). Turtle cortical sheets were perfused with normoxic (95%O(2)/5%CO(2)), anoxic (95%N(2)/5%CO(2)), high calcium (4 mM) and adenosine (200 microm) artificial cerebral spinal fluid (aCSF). No significant change in C(w) was observed under any of the above conditions. However, during hypo-osmotic aCSF perfusion C(w) decreased significantly, with the lowest value of 50+/-10.4 pF (P<0.05) occurring at 30 min. To visualize changes in gap junction permeability lucifer yellow was loaded into turtle neurons during normoxic, anoxic, 0 calcium, hypo-osmotic, cold shock, (+)-isoproterenol, nitric oxide donor S-nitoso-acetyl penicillamine, and 8-bromo-guanosine 3',5'-cyclic monophosphate aCSF perfusion. Dye propagation was only observed in 3 of 20 cold shock experiments (4 degrees C). We conclude that gap junctions are not involved in the acute reduction in G(w) previously observed during anoxia and that our results support the hypothesis that ion channel arrest is involved.

Identification of neural stem cells in the adult vertebrate brain

Neurogenesis continues into adult life in restricted germinal layers. The identification of the neural stem cells that give rise to these new neurons has important clinical implications and provides fundamental information to understand the origins of the new neurons. Work in adult birds and rodents yielded a surprising result: the neural stem cells appear to have characteristics of glia. In adult birds, the primary neuronal precursors are radial glia. In adult mammals, the primary neuronal precursors have properties of astrocytes. Radial glial cells have previously been shown to transform into astrocytes; both cell types are classically considered part of a committed astroglial lineage. Instead, we propose that neural stem cells are contained within this astroglial lineage. These findings in adult vertebrate brain, together with recent work in the developing mammalian cerebral cortex, force us to reexamine traditional concepts about the origin of neurons and glia in the central nervous system. In particular, neural stem cells possess a surprisingly elaborate structure, suggesting that in addition to their progenitor role, they have important structural and metabolic support functions. The very same cells that give birth to new neurons also seem to nurture their maturation and support their function.

Electrical coupling between glial cells in the rat retina

The strength of electrical coupling between retinal glial cells was quantified with simultaneous whole-cell current-clamp recordings from astrocyte-astrocyte, astrocyte-Müller cell, and Müller cell-Müller cell pairs in the acutely isolated rat retina. Experimental results were fit and space constants determined using a resistive model of the glial cell network that assumed a homogeneous two-dimensional glial syncytium. The effective space constant (the distance from the point of stimulation to where the voltage falls to 1/e) equaled 12.9, 6.2, and 3.7 microm, respectively for astrocyte-astrocyte, astrocyte-Müller cell, and Müller cell-Müller cell coupling. The addition of 1 mM Ba(2+) had little effect on network space constants, while 0.5 mM octanol shortened the space constants to 4.7, 4.4, and 2.6 microm for the three types of coupling. For a given distance separating cell pairs, the strength of coupling showed considerable variability. This variability in coupling strength was reproduced accurately by a second resistive model of the glial cell network (incorporating discrete astrocytes spaced at varying distances from each other), demonstrating that the variability was an intrinsic property of the glial cell network. Coupling between glial cells in the retina may permit the intercellular spread of ions and small molecules, including messengers mediating Ca(2+) wave propagation, but it is too weak to carry significant K(+) spatial buffer currents.

A unified hypothesis on the lineage of neural stem cells

For many years, it was assumed that neurons and glia in the central nervous system were produced from two distinct precursor pools that diverged early during embryonic development. This theory was partially based on the idea that neurogenesis and gliogenesis occurred during different periods of development, and that neurogenesis ceased perinatally. However, there is now abundant evidence that neural stem cells persist in the adult brain and support ongoing neurogenesis in restricted regions of the central nervous system. Surprisingly, these stem cells have the characteristics of fully differentiated glia. Neuroepithelial stem cells in the embryonic neural tube do not show glial characteristics, raising questions about the putative lineage from embryonic to adult stem cells. In the developing brain, radial glia have long been known to produce cortical astrocytes, but recent data indicate that radial glia might also divide asymmetrically to produce cortical neurons. Here we review these new developments and propose that the stem cells in the central nervous system are contained within the neuroepithelial --> radial glia --> astrocyte lineage.

Acute reduction in whole cell conductance in anoxic turtle brain

We tested the effect of anoxia, a "mimic" turtle artificial cerebrospinal fluid (aCSF) consisting of high Ca2+ and Mg2+ concentrations and low pH and adenosine perfusions, on whole cell conductance (G(w)) in turtle brain slices using a whole cell voltage-clamp technique. With EGTA in the recording electrode, anoxic or adenosine perfusions did not change Gw significantly (values range between 2.15 +/- 0.24 and 3.24 +/- 0.56 nS). However, perfusion with normoxic or anoxic mimic aCSF significantly decreased Gw. High [Ca2+] (4.0 or 7.8 mM) perfusions alone could reproduce the changes in Gw found with the mimic perfusions. With the removal of EGTA from the recording electrode, Gw decreased significantly during both anoxic and adenosine perfusions. The A1-receptor agonist N6-cyclopentyladenosine reduced Gw in a dose-dependent manner, whereas the A1-receptor specific antagonist 8-cyclopentyl-1,3-dipropylxanthine blocked both the adenosine- and anoxic-mediated changes in Gw. These data suggest a mechanism involving A1-receptor-mediated changes in intracellular [Ca2+] that result in acute changes in Gw with the onset of anoxia.

pH regulation and proton signalling by glial cells

The regulation of H+ in nervous systems is a function of several processes, including H+ buffering, intracellular H+ sequestering, CO2 diffusion, carbonic anhydrase activity and membrane transport of acid/base equivalents across the cell membrane. Glial cells participate in all these processes and therefore play a prominent role in shaping acid/base shifts in nervous systems. Apart from a homeostatic function of H(+)-regulating mechanisms, pH transients occur in all three compartments of nervous tissue, neurones, glial cells and extracellular spaces (ECS), in response to neuronal stimulation, to neurotransmitters and hormones as well as secondary to metabolic activity and ionic membrane transport. A pivotal role for H+ regulation and shaping these pH transients must be assigned to the electrogenic and reversible Na(+)-HCO3-membrane cotransport, which appears to be unique to glial cells in nervous systems. Activation of this cotransporter results in the release and uptake of base equivalents by glial cells, processes which are dependent on the glial membrane potential. Na+/H+ and Cl-/HCO3-exchange, and possibly other membrane carriers, accomplish the set of tools in both glial cells and neurones to regulate their intracellular pH. Due to the pH dependence of a great variety of processes, including ion channel gating and conductances, synaptic transmission, intercellular communication via gap junctions, metabolite exchange and neuronal excitability, rapid and local pH transients may have signalling character for the information processing in nervous tissue. The impact of H+ signalling under both physiological and pathophysiological conditions will be discussed for a variety of nervous system functions.

Octanol, a gap junction uncoupling agent, changes intracellular [H+] in rat astrocytes

Octanol rapidly closes gap junction channels but its mechanism of action is not known. Because intracellular [H+], pHi, also affects the conductance of gap junctions, we studied octanol's effects on pHi in cultured rat astrocytes, which are highly coupled cells. Octanol (1 mM) caused an acid shift in the pHi of 90% of rat hippocampal astrocytes which averaged -0.19 +/- 0.09 pH units in magnitude. In 58% of the cells tested, a biphasic change in pHi was seen; octanol produced an initial acidification lasting approximately 10 min that was followed by a persistent alkalinization. The related gap junction uncoupling agent, heptanol, had similar effects on pHi. Octanol-induced changes in pHi were similar in nominally HCO(3-)-free and HCO(3-)-containing solutions, although the rate of initial acidification was significantly greater in the presence of HCO3-. The initial acidification was inhibited in the presence of the stilbene DIDS, an inhibitor of Na+/HCO3- cotransport, indicating that octanol caused acidification by blocking this powerful acid extruder. The alkalinization was inhibited by amiloride which blocks the Na+/H+ exchanger (NHE), an acid extruder, suggesting that the alkaline shift induced by octanol was caused by stimulation of NHE. As expected, octanol's effects on astrocytic pHi were prevented by removal of external Na+, which blocks both Na+/HCO3- cotransport and NHE. Octanol had only small effects on intracellular Ca2+ (Ca2+i) in astrocytes. Hepatocytes which, like astrocytes, are strongly coupled to one another, showed no change in pHi with octanol application. Fluorescence recovery after photobleaching (FRAP) was used to study the effect of changes in astrocyte pHi on degree of coupling in hippocampal astrocytes. Coupling was decreased by intracellular acid shifts approximately -0.2 pH units in size. Octanol's effects on astrocyte pHi were complex but a prompt initial acidification was nearly always seen and could contribute to the uncoupling action of this drug in astrocytes. Because octanol uncouples hepatocytes without changing their pHi, this compound clearly can influence gap junctional conductance independent of changes in pHi.

The ependyma: a protective barrier between brain and cerebrospinal fluid

This review summarizes the current scientific literature concerning the ependymal lining of the cerebral ventricles of the brain with an emphasis on selective barrier function and protective roles for the common ependymal cell. Topics covered include the development, morphology, protein and enzyme expression including reactive changes, and pathology. Some cells lining the neural tube are committed at an early stage to becoming ependymal cells. They serve a secretory function and perhaps act as a cellular/axonal guidance system, particularly during fetal development. In the mature mammalian brain ependymal cells possess the structural and enzymatic characteristics necessary for scavenging and detoxifying a wide variety of substances in the CSF, thus forming a metabolic barrier at the brain-CSF interface.

Glial cells derived from aged mouse brain in culture display both mature and immature astrocytic phenotypes

In earlier studies, we established glial cell cultures derived from aged (18-month-old) mouse cerebral hemispheres (MACH) and have maintained them frozen at various passages. These cultures were characterized immunocytochemically and consist of: 5% oligodendrocytes (GalC+), 75% astrocytes-type 1 (GFAP+ only), 15% astrocytes-type 2 (GFAP+ + A2B5+), and 5% progenitor glial cells (A2B5+ only). In the present study, we isolated colonies from MACH passage 29 cultures and also colonies from MACH passage 19 transfected with the gene for SV40 large T antigen and further subcultured for 8 passages. Using double-staining immunocytochemistry, we found in non-transfected MACH passage 19 colonies consisting primarily of cells exhibiting only vimentin-positive staining and are considered to be immature glioblasts; colonies consisting primarily of cells exhibiting GFAP+ + vimentin+ which are considered to be astrocytes at an intermediate stage of maturation; and colonies consisting predominantly of cells exhibiting GFAP+ only which are considered to be mature astrocytes. In contrast, colonies isolated from transfected MACH cultures consisted primarily of vimentin+ cells. In conclusion, astrocytes in cultures derived from aged brain continue to be variable as they are during development. However, their response to the microenvironment may differ during development and during aging. Thus, the availability of clones of mature and immature astrocytes offers the opportunity to study neuron-glia interactions and the role of mature and immature astrocytes in neuronal aging and regeneration.

Reciprocal expression of cell-cell coupling and voltage-dependent Na current during embryogenesis of rat telencephalon

Using whole-cell patch-clamp techniques in situ (whole-tissue and tissue slices), we have studied two aspects of rat telencephalic cell development during the period of embryogenesis starting at E12. The first aspect was related to junctional coupling as revealed by low input resistance, intercellular dye spread and pharmacologic blockade. Coupling appeared to decrease with time, both in extent and occurrence. The second aspect dealt with cell excitability as revealed by voltage-dependent Na current (INa) expression. Immature action potentials and their underlying INaS were present in a small proportion of E12 cells. These currents were blocked 36% and 78% by 10(-7) M and 10(-6) M tetrodotoxin (TTX), respectively. From then onward, INaS got larger and more prevalent while no obvious changes in kinetics were observed. At E21, INaS were abolished by 10(-7) M TTX and channel density apparently was sufficient to support overshooting yet still immature action potentials.

Development of cell-cell coupling among cells of the oligodendrocyte lineage

The development of functional gap-junctional communication was studied in cells of the oligodendrocyte lineage. The presence of cell-cell coupling was determined by the passage of current between cells using a double whole-cell patch-clamp system or by injecting the low molecular weight dye Lucifer yellow into individual cells via a patch pepette and observing the diffusion of the dye into adjacent cells. The developmental stage of the cells under study was determined using antibodies to specific surface markers (04, 01, and 010) that characterize cells of sequential maturity along the oligodendrocyte lineage (Kuhlmann-Krieg et al., 1988; Sommer and Schachner, 1981; 1982). Both stages of precursor cells of this lineage, O4+ and O4-, almost never showed dye or electrical coupling, even though they were in close physical contact with other cells. The O1-positive oligodendrocytes with simple morphology were also noncoupled, with only few exceptions. In contrast, more than 40 percent of more mature, O10-positive oligodendrocytes showed cell-cell coupling detectable with both dye and current injection. Thus, the formation of gap junctions between cells of the oligodendrocyte lineage does not occur with, but some time after, the commitment of the cell to becoming an oligodendrocyte.

Radial glia give rise to perinodal processes

Nodes of Ranvier in the central nervous system in mammals are characterized by the presence of perinodal astrocytic processes. This study examines the association between processes of radial glia and the axolemma at nodes of Ranvier in the spinal cord of the mature axolotl, an animal in which radial glia represent a large portion of the total glial population. The radial glial cells have their cell bodies located close to the central canal. Those situated dorsal to the canal send long processes to the dorsal surface of the spinal cord. Along this trajectory these processes coalesce into large fascicles in the midline and form the dorsal median septum. Slender branches rise from the processes in these fascicles and extend into the adjacent white matter to terminate in close apposition to the axolemma at nodes of Ranvier. This arrangement provides an intracellular pathway extending from the perinodal region to the surface of the spinal cord. Radial glia ventral to the central canal give rise to processes that project to the glia limitans adjacent to the ventral spinal artery. These ventrally projecting processes appear to be more irregular in their branching pattern than their dorsal counterparts. Multiple slender processes are seen in close apposition to the nodal axolemma of myelinated axons in the ventral white commissure, again providing an intracellular pathway that runs from the perinodal region to the cord surface. In one instance a radial glial process was observed to occupy a pocket formed by the invagination of the nodal axolemma. The axonal cytoplasm adjacent to the invagination contained a variety of organelles, e.g. multivesicular bodies, vesicles and endoplasmic reticulum, suggesting that this relationship between the radial glial process and the axon is more than a passive interaction. These observations are consistent with the view that processes of radial glial cells may regulate the extracellular environment adjacent to the nodal axolemma, and/or play an active role in the maintenance of the nodal membrane. The existence of perinodally-directed processes of radial glial cells in the salamander indicates that this axo-glial specialization reflects an important functional interaction preserved across a large segment of the phylogenetic scale.

Mechanism of neurogenesis in adult avian brain

Adult neurogenesis in birds offers unique opportunities to study basic questions addressing the birth, migration and differentiation of neurons. Neurons in adult canaries originate from discrete proliferative regions on the walls of the lateral ventricles. They migrate away from their site of birth, initially at high rates, along the processes of radical cells. The rates of dispersal diminish as the young neurons invade regions devoid of radial fibers, probably under the guidance of other cues. The discrete sites of birth in the ventricular zone generate neurons that end up differentiating throughout the telencephalon. New neurons may become interneurons or projection neurons; the latter connect two song control nuclei between neostriatum and archistriatum. Radial cells, that in mammals disappear as neurogenesis comes to an end, persist in the adult avian brain. The presence of radial cells may be key to adult neurogenesis. Not only do they serve as guides for initial dispersal, they also divide and may be the progenitors of new neurons.

Proliferation "hot spots" in adult avian ventricular zone reveal radial cell division

Neurogenesis in the adult avian brain is restricted to the telencephalon. New neurons originate in the ventricular zone (VZ) from cells that have not been identified. We mapped the position of [3H]thymidine-labeled cells in the walls of the ventricles of the adult canary brain. Labeled VZ cells were restricted to the telencephalon (lateral ventricles) and concentrated in "hot spots". The coincidence of these hot spots with regions rich in radial cells suggested that radial cells may be the cells undergoing mitosis. We used smears prepared from fragments of the VZ containing the hot spots to show directly that radial cells accumulate [3H]thymidine. In addition, grain counts at different survival times demonstrated that these cells divide. Hot spots of VZ cell division also coincided with sites of neuronal origin. We suggest that radial cell division may give rise to new neurons.

Glutamate- and aspartate-induced extracellular potassium and calcium shifts and their relation to those of kainate, quisqualate and N-methyl-D-aspartate in the isolated turtle cerebellum

Ion-selective microelectrodes can be used to evaluate the characteristics and laminar distribution of excitatory amino acid agonist-induced K+ and Ca2+ shifts in the extracellular environment of brain cells. This report describes the pattern of K+ increases and Ca2+ decreases elicited by glutamate and aspartate at 100 microns intervals in the isolated turtle cerebellum. These responses were compared to ion shifts evoked by kainate, quisqualate and N-methyl-D-aspartate. Glutamate and aspartate produced indistinguishable laminar patterns of ion shifts, with the greatest [K+]o and [Ca2+]o shifts in the granular layer. The average maximum granular and molecular layer increases in [K+]o were, respectively, 130% and 24% larger than the increase in the Purkinje cell layer. Kainate-induced increases in [K+]o also followed this granular greater than molecular greater than Purkinje cell layer pattern; however, the corresponding [Ca2+]o decreases were smaller and more variable. Quisqualate-evoked ion shifts in the molecular layer closely mimicked the shape of glutamate- and aspartate-induced responses. In the granular layer, however, quisqualate caused little ion change during iontophoresis followed by large [K+]o and [Ca+]o shifts after the end of the pulse. The minimal ion shifts induced during quisqualate application in the granular layer gave this agonist the distinction of being the only agent tested to have its greatest direct effect in the molecular layer. N-Methyl-D-aspartate caused large, two-phase [K+]o and [Ca2+]o shifts in the granular layer, only small [K+]o rises in the Purkinje cell and ventral molecular layers, and no response in the dorsal molecular layer. The lack of similarity between glutamate- and aspartate-induced ion shifts in the granular layer and those of any one agonist demonstrate the mixed agonist action of glutamate and aspartate in the cerebellum. These studies provide new information about the dynamics of excitatory amino acid receptor activation that is complementary to autoradiographic receptor mapping data and to single cell electrophysiological studies.

Electrical coupling, without dye coupling, between mammalian astrocytes and oligodendrocytes in cell culture

Evidence of electrical and dye coupling between oligodendrocytes and astrocytes was sought in cultures of mouse spinal cord. Cell identity was verified using cell specific antigenic markers. In most experiments current was injected into oligodendrocytes while recording voltage in nearby astrocytes. Nine of 17 oligodendrocyte-astrocyte cell pairs showed weak electrical coupling; the average estimated coupling ratio was 0.03 +/- 0.06 (cf. 0.11 for oligodendrocyte-oligodendrocyte and 0.44 for astrocyte-astrocyte pairs; Kettenmann and Ransom: Glia, 1: 64-73, 1988). Application of 0.5 mM BaCl2 or 44.6 mM CsCl depolarized astrocytes and oligodendrocytes and was estimated to increase the coupling ratio between these cells 3-5-fold; these effects were rapid in onset and completely reversible. In 5 of 7 cases, oligodendrocyte-astrocyte pairs that appeared uncoupled in normal solution exhibited coupling during Ba++ or Cs+ exposure. The actions of these cations are believed to be mediated by blockade of glial K+ channels. Depolarization, per se, as induced by increasing [K+]o, did not increase coupling ratio. The fluorescent dye lucifer yellow (LY) was injected into 10 oligodendrocytes, 8 of which were electrically coupled to nearby astrocytes, and never passed into astrocytes in detectable quantities. Likewise, astrocytes injected with LY stained other astrocytes, but never oligodendrocytes. These findings document the presence of weak electrical coupling between astrocytes and oligodendrocytes, in the absence of dye coupling. Weak coupling of this sort could subserve metabolic interactions between these cells mediated by the passage of small but important molecules such as cyclic AMP, but would not allow strong electrical interactions. If such coupling among glial cells is widespread, it would constitute a "metabolic syncytium" that could serve to coordinate glial behavior.

Intracellular pH shifts capable of uncoupling cultured oligodendrocytes are seen only in low HCO3- solution

Electrical coupling between cultured mouse oligodendrocytes was transiently blocked when pHi was decreased below about 6.5 using the NH4+ prepulse method. This uncoupling could, however, only be achieved if the dominant pHi regulating mechanism in these cells, the Na+/HCO3- cotransporter, was blocked by lowering bath [HCO3-]. Under this condition, an NH4+ prepulse caused pHi to decrease toward the passive distribution for H+ (i.e., about pH 6.2). In the presence of normal bath [HCO3-] an NH4+ prepulse did not decrease pHi below 6.5 even when the second pHi regulating mechanism, the Na+/H+ exchanger, was blocked by amiloride, and consequently oligodendrocytes could not be uncoupled. Increasing CO2, which uncouples glial cells in situ (Connors et al: J. Neurosci. 4:1324-1330, 1984), did not uncouple cultured oligodendrocytes in the presence of normal bath [HCO3-], but did cause uncoupling in low [HCO3-] solution. These results indicate that electrical coupling between cultured oligodendrocytes is sensitive to pHi; in normal bath [HCO3-], however, the pHi regulation of these cells is so effective that standard techniques for intracellular acidification are unable to lower pHi to levels which cause the closure of oligodendrocyte gap junctions.

Relations between slow extracellular potential changes, glial potassium buffering, and electrolyte and cellular volume changes during neuronal hyperactivity in cat brain

The aim of this investigation is to estimate the contribution of spatial glial K+ buffer currents to extracellular K+ homeostasis during enhanced neuronal activity. Neuronal hyperactivity was induced by electrical stimulation of the cortical surface or the ventrobasal thalamic nuclei of cats (5-50 Hz, 0.1-0.2 ms, two to three times threshold stimulation intensity, 5-20 s). The accompanying slow field potential changes were recorded simultaneously across the grey matter with vertical assemblies of eight micropipettes glued 300 microns apart. Using the Poisson equation, the amplitudes of the underlying current sources and sinks were calculated. The current source densities depended on the depth of recording, frequency, strength, and duration of the stimulation. Current sinks, corresponding to a removal of 0.1-0.5 mmoles of monovalent cations per liter of brain tissue and second from the extracellular space, were observed in middle cortical layers, whereas sources appeared at superficial and deeper sites. These sinks and sources might represent K+ moved across glial membranes by spatial buffer currents. The consequences of glial buffer currents of this magnitude were investigated with model calculations. It turned out that measurements of electrolyte and volume changes of the extracellular space (Dietzel et al. Exp. Brain Res. 40:432-439, 1980; Exp. Brain Res. 46:73-84, 1982) could only partially be explained by spatial buffer currents of this magnitude. Comparison of the calculated values with intracellular measurements in neurons and glial cells (Coles et al. Ann. N.Y. Acad. Sci. 481:303-317, 1986; Ballanyi et al. J. Physiol. 382:159-174, 1987) suggests that spatial buffering combines with an approximately equimolar KCl transport and, depending on the preparation, also K+/Na+-exchange across glial membranes.

A quantitative analysis of glial cell coupling in the retina of the axolotl (Ambystoma mexicanum)

The strength of gap junctional coupling of radial glial cells (Müller cells) in the isolated axolotl retina was assessed by monitoring the spread of dye between cells, and by injecting current into one cell and recording the voltage response in surrounding cells. Dye injected into one Müller cell spread to surrounding Müller cells, and could be detected up to 130 micron away, i.e. over 4 times the mean Müller cell spacing of 30 micron. Injecting 1 nA of current into a Müller cell evoked responses of 7 mV in that cell, 1 mV in next neighbour cells, and 0.2 mV in cells at 60 micron distance. Analysis of these data indicates an electrical space constant for the Müller cell network of 15 micron, and predicts that isolated cells should have a resistance of 11.4 M omega. Müller cells isolated by papain dissociation of the retina were found, by whole-cell patch-clamping, to have a mean resistance of 12.4 M omega. These results on lateral coupling are combined with data showing that over 90% of the Müller cell potassium conductance is in the vitreal endfoot of these cells to provide a fairly complete electrical description of the radial glial cell network in the retina. Gap junctional coupling of Müller cells increases by 60% the 'spatial buffering' that these glial cells can carry out to reduce localized rises in extracellular potassium concentration. The location of the majority of the Müller cell potassium conductance in the cell endfoot ensures that laterally buffered K+ is deposited in the vitreous, rather than depolarizing surrounding retinal neurones.

Gap junctions in goldfish preoptic ependyma: regional variation in cellular differentiation

Ependyma adjacent to the goldfish preoptic neurosecretory nucleus was examined with transmission electron microscopy. Ependymal cells adjoining the rostroventral end of the nucleus were spindle-shaped with their long axes perpendicular to the ventricular surface. Gap junctions and desmosomes were common near the apical (ventricular) ends of these cells, and less frequent laterally in the ependymal layer. Ependymal cells in more caudodorsal preoptic regions (adjacent to large neurosecretory cells) were progressively more pleomorphic. The frequent occurrence of apparently internalized gap junctions and of gap junction fragments enclosed within lysosome-like organelles indicated extensive turnover of these junctions, or uncoupling. Ependymal cells in the caudodorsal region formed gap junctions on their lateral and basal (abluminal) surfaces with glial processes containing bundles of intermediate filaments. Subependymally, these processes (presumptive radial glia) were parallel to one another and coupled together by gap junctions. Neurites containing dense core vesicles occasionally invaginated into ependymal cells in the caudal region, but did not appear to form gap junctions. Previous observations indicate continuing maturation and growth of the goldfish preoptic area with neurosecretory cell formation rostroventrally and a rostroventral to caudodorsal gradient of maturation. The present findings suggest a parallel and related gradient in preoptic ependyma. Ependymal cell differentiation possibly involves loss of gap junctions, and radial migration or differentiation into underlying neurons and glia.