Structure and Functions of Gap Junctions and Their Constituent Connexins in the Mammalian CNS.

Numerous data obtained in the last 20 years indicate that all parts of the mature central nervous system, from the retina and olfactory bulb to the spinal cord and brain, contain cells connected by gap junctions (GJs). The morphological basis of the GJs is a group of joined membrane hemichannels called connexons, the subunit of each connexon is the protein connexin. In the central nervous system, connexins show specificity and certain types of them are expressed either in neurons or in glial cells. Connexins and GJs of neurons, combining certain types of inhibitory hippocampal and neocortical neuronal ensembles, provide synchronization of local impulse and rhythmic activity, thalamocortical conduction, control of excitatory connections, which reflects their important role in the processes of perception, concentration of attention and consolidation of memory, both on the cellular and at the system level. Connexins of glial cells are ubiquitously expressed in the brain, and the GJs formed by them provide molecular signaling and metabolic cooperation and play a certain role in the processes of neuronal migration during brain development, myelination, tissue homeostasis, and apoptosis. At the same time, mutations in the genes of glial connexins, as well as a deficiency of these proteins, are associated with such diseases as congenital neuropathies, hearing loss, skin diseases, and brain tumors. This review summarizes the existing data of numerous molecular, electrophysiological, pharmacological, and morphological studies aimed at progress in the study of the physiological and pathophysiological significance of glial and neuronal connexins and GJs for the central nervous system.

Hyperglycemia reduces functional expression of astrocytic Kir4.1 channels and glial glutamate uptake.

Diabetics are at risk for a number of serious health complications including an increased incidence of epilepsy and poorer recovery after ischemic stroke. Astrocytes play a critical role in protecting neurons by maintaining extracellular homeostasis and preventing neurotoxicity through glutamate uptake and potassium buffering. These functions are aided by the presence of potassium channels, such as Kir4.1 inwardly rectifying potassium channels, in the membranes of astrocytic glial cells. The purpose of the present study was to determine if hyperglycemia alters Kir4.1 potassium channel expression and homeostatic functions of astrocytes. We used q-PCR, Western blot, patch-clamp electrophysiology studying voltage and potassium step responses and a colorimetric glutamate clearance assay to assess Kir4.1 channel levels and homeostatic functions of rat astrocytes grown in normal and high glucose conditions. We found that astrocytes grown in high glucose (25 mM) had an approximately 50% reduction in Kir4.1 mRNA and protein expression as compared with those grown in normal glucose (5mM). These reductions occurred within 4-7 days of exposure to hyperglycemia, whereas reversal occurred between 7 and 14 days after return to normal glucose. The decrease in functional Kir channels in the astrocytic membrane was confirmed using barium to block Kir channels. In the presence of 100-µM barium, the currents recorded from astrocytes in response to voltage steps were reduced by 45%. Furthermore, inward currents induced by stepping extracellular [K(+)]o from 3 to 10mM (reflecting potassium uptake) were 50% reduced in astrocytes grown in high glucose. In addition, glutamate clearance by astrocytes grown in high glucose was significantly impaired. Taken together, our results suggest that down-regulation of astrocytic Kir4.1 channels by elevated glucose may contribute to the underlying pathophysiology of diabetes-induced CNS disorders and contribute to the poor prognosis after stroke.

Transport Reversal during Heteroexchange: A Kinetic Study.

It is known that secondary transporters, which utilize transmembrane ionic gradients to drive their substrates up a concentration gradient, can reverse the uptake and instead release their substrates. Unfortunately, the Michaelis-Menten kinetic scheme, which is popular in transporter studies, does not include transporter reversal, and it completely neglects the possibility of equilibrium between the substrate concentrations on both sides of the membrane. We have developed a complex two-substrate kinetic model that includes transport reversal. This model allows us to construct analytical formulas allowing the calculation of a "heteroexchange" and "transacceleration" using standard Michaelis coefficients for respective substrates. This approach can help to understand how glial and other cells accumulate substrates without synthesis and are able to release such substrates and gliotransmitters.

L-DOPA Uptake in Astrocytic Endfeet Enwrapping Blood Vessels in Rat Brain.

Astrocyte endfeet surround brain blood vessels and can play a role in the delivery of therapeutic drugs for Parkinson's disease. However, there is no previous evidence of the presence of LAT transporter for L-DOPA in brain astrocytes except in culture. Using systemic L-DOPA administration and a combination of patch clamp, histochemistry and confocal microscopy we found that L-DOPA is accumulated mainly in astrocyte cell bodies, astrocytic endfeet surrounding blood vessels, and pericytes. In brain slices: (1) astrocytes were exposed to ASP(+), a fluorescent monoamine analog of MPP(+); (2) ASP(+) taken up by astrocytes was colocalized with L-DOPA fluorescence in (3) glial somata and in the endfeet attached to blood vessels; (4) these astrocytes have an electrogenic transporter current elicited by ASP(+), but intriguingly not by L-DOPA, suggesting a different pathway for monoamines and L-DOPA via astrocytic membrane. (5) The pattern of monoamine oxidase (MAO type B) allocation in pericytes and astrocytic endfeet was similar to that of L-DOPA accumulation. We conclude that astrocytes control L-DOPA uptake and metabolism and, therefore, may play a key role in regulating brain dopamine level during dopamine-associated diseases. These data also suggest that different transporter mechanisms may exist for monoamines and L-DOPA.

Downregulation of Kir4.1 inward rectifying potassium channel subunits by RNAi impairs potassium transfer and glutamate uptake by cultured cortical astrocytes.

Glial cell-mediated potassium and glutamate homeostases play important roles in the regulation of neuronal excitability. Diminished potassium and glutamate buffering capabilities of astrocytes result in hyperexcitability of neurons and abnormal synaptic transmission. The role of the different K+ channels in maintaining the membrane potential and buffering capabilities of cortical astrocytes has not yet been definitively determined due to the lack of specific K+ channel blockers. The purpose of the present study was to assess the role of the inward-rectifying K+ channel subunit Kir4.1 on potassium fluxes, glutamate uptake and membrane potential in cultured rat cortical astrocytes using RNAi, whole-cell patch clamp and a colorimetric assay. The membrane potentials of control cortical astrocytes had a bimodal distribution with peaks at -68 and -41 mV. This distribution became unimodal after knockdown of Kir4.1, with the mean membrane potential being shifted in the depolarizing direction (peak at -45 mV). The ability of Kir4.1-suppressed cells to mediate transmembrane potassium flow, as measured by the current response to voltage ramps or sequential application of different extracellular [K+], was dramatically impaired. In addition, glutamate uptake was inhibited by knock-down of Kir4.1-containing channels by RNA interference as well as by blockade of Kir channels with barium (100 microM). Together, these data indicate that Kir4.1 channels are primarily responsible for significant hyperpolarization of cortical astrocytes and are likely to play a major role in potassium buffering. Significant inhibition of glutamate clearance in astrocytes with knock-down of Kir4.1 highlights the role of membrane hyperpolarization in this process.

Tandem-pore domain potassium channels are functionally expressed in retinal (Müller) glial cells.

Tandem-pore domain (2P-domain) K+-channels regulate neuronal excitability, but their function in glia, particularly, in retinal glial cells, is unclear. We have previously demonstrated the immunocytochemical localization of the 2P-domain K+ channels TASK-1 and TASK-2 in retinal Müller glial cells of amphibians. The purpose of the present study was to determine whether these channels were functional, by employing whole-cell recording from frog and mammalian (guinea pig, rat and mouse) Müller cells and confocal microscopy to monitor swelling in rat Müller cells. TASK-like immunolabel was localized in these cells. The currents mediated by 2P-domain channels were studied in isolation after blocking Kir, K(A), K(D), and BK channels. The remaining cell conductance was mostly outward and was depressed by acid pH, bupivacaine, methanandamide, quinine, and clofilium, and activated by alkaline pH in a manner consistent with that described for TASK channels. Arachidonic acid (an activator of TREK channels) had no effect on this conductance. Blockade of the conductance with bupivacaine depolarized the Müller cell membrane potential by about 50%. In slices of the rat retina, adenosine inhibited osmotic glial cell swelling via activation of A1 receptors and subsequent opening of 2P-domain K+ channels. The swelling was strongly increased by clofilium and quinine (inhibitors of 2P-domain K+ channels). These data suggest that 2P-domain K+ channels are involved in homeostasis of glial cell volume, in activity-dependent spatial K+ buffering and may play a role in maintenance of a hyperpolarized membrane potential especially in conditions where Kir channels are blocked or downregulated.

Characterization of acid-sensitive ion channels in freshly isolated rat brain neurons.

Transient proton-activated currents induced by rapid shifts of the extracellular pH from 7.4 to < or =6.8 were recorded in different neurons freshly isolated from rat brain (hypoglossal motoneurons, cerebellar Purkinje cells, striatal giant cholinergic interneurons, hippocampal interneurons, CA1 pyramidal neurons and cortical pyramidal neurons) using whole-cell patch clamp technique. Responses of hippocampal CA1 pyramidal neurons were weak (100-300 pA) in contrast to other types of neurons (1-3 nA). Sensitivity of neurons to rapid acidification varied from pH(50) 6.4 in hypoglossal motoneurons to 4.9 in hippocampal interneurons. Proton-activated currents were blocked by amiloride (IC(50) varied from 3.6 to 9.5 microM). Reversal potential of the currents was close to E(Na), indicating that the currents are carried by sodium ions. The data obtained suggest that the proton-activated currents in the neurons studied are mediated by acid-sensitive ion channels. Strong acidification (pH<4) induced biphasic responses in all neuron types: the transient current was followed by a pronounced sustained one. Sustained current was not blocked by amiloride and exhibited low selectivity for sodium and cesium ions. Slow acidification from pH 7.4 to 6.5 did not induce detectable whole-cell currents. At pH 6.5, most of the channels are desensitized and responses to fast pH shifts from this initial level are decreased at least 10 times. This suggests that slow acidification which is well known to accompany some pathological states should rather desensitize than activate acid-sensitive ion channels and depress their function. Our results provide evidence for a widespread and neuron-specific distribution of acid-sensitive ion channels in the brain. The large amplitudes and transient character of currents mediated by these channels suggest that they could contribute to fast neuronal signaling processes.

Kir6.1 is the principal pore-forming subunit of astrocyte but not neuronal plasma membrane K-ATP channels.

ATP-sensitive potassium channels (K-ATP channels) directly couple the energy state of a cell to its excitability, are activated by hypoxia, and have been suggested to protect neurons during disturbances of energy metabolism such as transient ischemic attacks or stroke. Molecular studies have demonstrated that functional K-ATP channels are octameric protein complexes, consisting of four sulfonylurea receptor proteins and four pore-forming subunits which are members of the Kir6 family of inwardly rectifying potassium channels. Here we show, using specific antibodies against the two known pore-forming subunits (Kir6.1 and Kir6.2) of K-ATP channels, that only Kir6.1 and not Kir6.2 subunits are expressed in astrocytes. In addition to a minority of neurons, Kir6.1 protein is present on hippocampal, cortical, and cerebellar astrocytes, tanycytes, and Bergmann glial cells. We also provide ultrastructural evidence that Kir6.1 immunoreactivity is primarily localized to distal perisynaptic and peridendritic astrocyte plasma membrane processes, and we confirm the presence of functional K-ATP channels in Bergmann glial cells by slice-patch-clamp experiments. The identification of Kir6.1 as the principal pore-forming subunit of plasma membrane K-ATP channels in astrocytes suggests that these glial K-ATP channels act in synergy with neuronal Kir6.2-mediated K-ATP channels during metabolic challenges in the brain.

Kir subfamily in frog retina: specific spatial distribution of Kir 6.1 in glial (Müller) cells.

We show by immunocytochemistry in frog retina that most members of the Kir subfamily are expressed in specific neuronal compartments. However, Kir 6.1, the pore-forming subunit of K(ATP) channels, is expressed exclusively in glial Müller cells. Müller cell endfeet display strong Kir 6.1 immunolabel throughout the retina, whereas the somata are labeled only in the retinal periphery. This spatial pattern is similar to that of Kir 4.1, of the ratio of inward to outward K+ currents, and of spermine/spermidine immunoreactivity. We suggest that the co-expression of Kir 4.1 and Kir 6.1 subunits may enable the cells to maintain their high K+ conductance and hyperpolarized membrane potentials both at high ATP levels (Kir 4.1) and during ATP deficiency (Kir 6.1).

Farnesol modulates membrane currents in human retinal glial cells.

Farnesol, a C(15) natural isoprenoid, exerts complex modulating effects on the membrane permeability of human retinal glial (Müller) cells. Several glial cationic currents were examined. At low micromolar concentrations, farnesol reduced the amplitudes of all fast and depolarization-activated membrane currents expressed by Müller cells, that is, currents through 1) transient low-voltage-activated (LVA; IC(50) = 2.2 microM), 2) sustained high-voltage-activated Ca(2+) channels (HVA; IC(50) = 1.2 microM), 3) fast Na(+) channels (IC(50) = 9.0 microM), and 4) transient (A-type) K(+) channels (IC(50) = 4.7 microM). Furthermore, farnesol shifted the activation of LVA and HVA currents to more depolarized potentials by 21.3 +/- 7.4 mV and 8.3 +/- 4.5 mV, respectively. On the other hand, neither inwardly rectifying nor iberiotoxin-sensitive calcium-activated K(+) currents were affected by farnesol. Therefore, farnesol is assumed to be a biologically active substance that regulates ion channel activity in the glial cell membrane. Depressing rapid changes of the membrane potential and supporting a stable hyperpolarized status of the glial cells may enhance the efficiency of crucial glial functions such as extracellular K(+) clearance and neurotransmitter uptake.

Müller glial cells in anuran retina.

Whereas in the brain, the activity of the neurons is supported by several types of glial cells such as astrocytes, oligodendrocytes, and ependymal cells, the retina (evolving from the brain during ontogenesis) contains only one type of macroglial cell, the Müller (radial glial) cells, in most vertebrates including the anurans. These cells span the entire thickness of the tissue, and thereby contact and ensheath virtually every type of neuronal cell body and process. This intimate topographical relationship is reflected by a multitude of functional interactions between retinal neurons and Müller glial cells. Müller cells are the principal stores of retinal glycogen, and are thought to fuel retinal neurons with substrate (lactate/pyruvate) for their oxidative metabolism. Furthermore, Müller cells are involved in the control and homeostasis of many constituents of the extracellular space, such as potassium and perhaps other ions, signaling molecules, and of the extracellular pH. They also seem to play important roles in recycling mechanisms of photopigment molecules and neurotransmitter molecules such as glutamate and GABA. By containing the main retinal stores of glutathione, Müller cells may protect retinal neurons against free radicals. Moreover, Müller cells express receptors for many neuroactive substances, and may also release such substances to their neighbouring neurons. Thus, Müller cells exert many functions crucial for signal processing in the normal retina. Moreover, Müller cells change their properties in cases of retinal disease and injury, and may either support the survival of neuronal cells or accelerate the progress of neuronal degeneration.

Spatial distribution of spermine/spermidine content and K(+)-current rectification in frog retinal glial (Müller) cells.

Previous studies in retinal glial (Müller) cells have suggested that (1) the dominant membrane currents are mediated by K(+) inward-rectifier (Kir) channels (Newman and Reichenbach, Trends Neurosci 19:307-312, 1996), and (2) rectification of these Kir channels is due largely to a block of outward currents by endogenous polyamines such as spermine/spermidine (SPM/SPD) (Lopatin et al., Nature 372:366-369, 1994). In frog Müller cells, the degree of rectification of Kir-mediated currents is significantly higher in the endfoot than in the somatic membrane (Skatchkov et al., Glia 27:171-181, 1999). This article shows that in these cells there is a topographical correlation between the local cytoplasmic SPM/SPD immunoreactivity and the ratio of inward to outward K(+) currents through the surrounding membrane area. Throughout the retina, Müller cell endfeet display a high SPM/SPD immunolabel (assessed by densitometry) and a large inward rectification of K(+) currents, as measured by the ratio of inward to outward current produced by step changes in [K(+)](o). In the retinal periphery, Müller cell somata are characterized by roughly one-half of the SPM/SPD immunoreactivity and K(+)-current rectification as the corresponding endfeet. In the retinal center, Müller cell somata are virtually devoid of both SPM/SPD immunolabel and K(+)-current inward rectification. Comparing one region of the retina with another, we find an exponential correlation between the local K(+) rectification and the local SPM/SPD content. This finding suggests that the degree of inward rectification in a given membrane area is determined by the local cytoplasmic polyamine concentration.

Ca(2+) channel-mediated currents in retinal glial (Müller) cells of the toad (Bufo marinus).

Whole-cell voltage-clamp recordings were used to detect voltage-gated Ca(2+) channels in freshly isolated retinal glial (Müller) cells of the toad (Bufo marinus). Using Ca(2+) ions (2 mM) as charge carriers (in the presence of 1 mM Mg(2+)), no inwardly directed currents could be observed during the application of depolarizing voltage steps. However, after omitting the divalent cations from the bath solution, large-amplitude inwardly directed currents were evoked that were carried by Na(+) ions, and were mediated by at least two different kinds of Ca(2+) channels, transient low voltage-activated (LVA) channels and sustained high voltage-activated (HVA) channels. While the LVA currents activated at potentials positive to -90 mV and peaked at -40 mV, the HVA currents activated positive to -60 mV and peaked at -20 mV. It is concluded that Müller glial cells of the toad express distinct types of voltage-gated Ca(2+) channels that may be activated, under certain conditions, close to physiological membrane potentials.

Potassium buffering by Müller cells isolated from the center and periphery of the frog retina.

Müller (radial glial) cells span the retina from the outer to the inner limiting membranes. They are the only glial cells found in the amphibian retina. The thickness of the frog (Rana pipiens) retina decreases by a factor of about four from the center to the periphery. Thus, Müller cells were isolated, by enzymatic dissociation, with stalk lengths from 20 to 140 microm. Their ability to transfer K(+) via the stalk between soma and endfoot was studied. Membrane currents were recorded using the whole-cell voltage-clamp technique with the pipette sealed to either the endfoot or the soma. Inward (I(KIN)) or outward (I(KO)) currents were elicited by rapid increases (3 to 10 mM) or decreases (3 to 1 mM) of the extracellular K(+) concentration ([K(+)](o)) either by local application (close or distant to the recording pipette) or around the entire cell (whole cell perfusion). For the long central cells, the ratio I(KIN)/I(KO) was 4.6 +/- 0.6 SE (n = 9) at the endfoot and 1.7 +/- 0.1 SE (n = 8) at the soma. In cells from the retinal periphery, the ratio I(KIN)/I(KO) was higher, 7.0 +/- 0.27 (n = 8) at the endfoot and 3.2 +/- 0.1 (n = 10) at the soma. The results suggest that there is less inward rectification in the somatic than in the endfoot membrane. As expected from previous studies, the sensitivity of the cells to K(+) was higher at the endfoot than at the soma. The amplitude of I(KIN) at the endfoot compared to the soma was about 8-fold for the long central cells but only about 1.5-fold for the short peripheral cells. Currents spread readily from endfoot to soma in the peripheral cells. In the long central Müller cells the soma and endfoot appeared electrotonically isolated. The "functional length constant", lambda, of cell stalk processes was about 70 microm. The relative decrement of large inward currents was stronger than that of smaller outward currents; this difference ("artificial rectification") is explained by a simple model, where larger currents (inward) are attenuated more than smaller (outward) currents. The data support the hypothesis that in the retinal periphery, Müller cells provide extensive spatial K(+) buffering from both plexiform layers into the vitreous body. In the central retina, however, such currents are limited within a short (interlaminar) range.

The activity of a transient potassium current in retinal glial (Müller) cells depends on extracellular calcium.

The modulating effects of varying extracellular concentrations of Ca2+ ([Ca2+]e) and of other divalent cations on the fast transient (A-type) K+ current (I(A)) of freshly isolated Muller glial cells from rabbit and human retinae were studied with the whole-cell patch-clamp method. The I(A) of Miller cells was voltage-independently blocked by extracellular 4-aminopyridine (4AP) with a 50 % reduction achieved at 0.94 mM 4AP. The I(A) amplitude was elevated by increased extracellular [K+]. Elevation of the [Ca2+]e had three effects on the glial I(A): (i) it concentration-dependently shifted both the activation and inactivation curves towards less negative membrane potentials, (ii) it increased the peak current amplitude, and (iii) it slowed down the activation and inactivation kinetics. Particularly at depolarized membrane potentials, the I(A) was enlarged and broadened when the [Ca2+]e was increased. Various divalent cations also exerted these effects, although at different concentrations. While Zn2+, Cd2+, Cu2+ and Pb2+ modulated the I(A) in the micromolar range, Mg2+ and Ba2+ had effects in the millimolar range. Extracellular acidification produced a positive shift in the voltage dependence of I(A) gating. However, alterations of the extracellular pH did not abolish the Ca2+ effects on I(A); this indicates that protons and Ca2+ ions mediate their effects on glial K(A) channels by different mechanisms or binding sites, respectively. Physiological (i.e., activity-dependent) changes of the extracellular concentration of divalent cations and of the extracellular pH should influence the retinal excitability via modulation of glial K+ currents. The activation of glial I(A) by divalent cations at depolarized voltages supports a repolarization and, therefore, the maintainance of a hyperpolarized glial membrane potential during periods of increased neuronal activity.

Alterations of potassium channel activity in retinal Müller glial cells induced by arachidonic acid.

Arachidonic acid, which is thought to be involved in pathogenetic mechanisms of the central nervous system, has been shown previously to modulate neuronal ion channels and the glutamate uptake carrier of retinal glial (Müller) cells. We have used various configurations of the patch-clamp technique to determine the effects of arachidonic acid on the K+ currents of freshly isolated Müller glial cells from rabbit and human. Arachidonic acid reduced the peak amplitude of the transient (A-type) outward K+ currents in a dose-dependent and reversible manner, with a 50% reduction achieved by 4.1 microM arachidonic acid. The inward rectifier-mediated currents remained unchanged after arachidonic acid application. The amplitude of the Ca(2+)-activated K+ outward currents (KCa), which were blocked by 1 mM tetraethylammonium chloride and 40 nM iberiotoxin, respectively, was dose-dependently elevated by bath application of arachidonic acid. The activation curve of the KCa currents shifted towards more negative membrane potentials. Furthermore, arachidonic acid was found to suppress inwardly directed Na+ currents. In cell-attached recordings with 3 mM K+ in the bath and 130 mM K+ in the pipette, the KCa channels of rabbit Müller cells displayed a linear current-voltage relation, with a mean slope conductance of 102 pS. In excised patches, the slope conductance was 220 pS (150 mM K+i/130 mM K+o). The opening probability of the KCa channels increased during membrane depolarization and during elevation of the free Ca2+ concentration at the intracellular face of the membrane patches. Bath application of arachidonic acid caused a reversible increase of the single-channel opening probability, as well as an increase of the number of open channels. Arachidonic acid did not affect the single-channel conductance. Since arachidonic acid also stimulates the KCa channel activity in excised patches, the action of arachidonic acid is assumed to be independent of changes of the intracellular calcium concentration. Our results demonstrate that arachidonic acid exerts specific effects on distinct types of K+ channels in retinal glial, cells. In pathological cases, elevated arachidonic acid levels may contribute to prolonged Müller cell depolarizations, and to the initiation of reactive glial cell proliferation.

Spermine/spermidine is expressed by retinal glial (Müller) cells and controls distinct K+ channels of their membrane.

There is recent evidence that polyamines such as spermine (spm) and spermidine (spd) may act as endogenous modulators of the activity of inwardly rectifying K+ channels. This type of K+ channels is abundantly expressed by retinal glial (Müller) cells where they are involved in important glial cell functions such as the clearance of excess extracellular K+ ions. This prompted us to study the following questions, i) do mammalian Müller cells contain endogenous spm/spd?; ii) do Müller cells possess the enzymes (e.g., ornithine decarboxylase, ODC) necessary to produce spm/spd?; and iii) does application of exogenous spm/spd exert specific effects onto inwardly rectifying K+ channels of Müller cells? Immunocytochemical studies were performed on histological sections of guinea-pig, rabbit, porcine, and human retinae, and on enzymatically dissociated Müller cells. Whole-cell and patch-clamp recordings were performed on enzymatically dissociated porcine and guinea-pig Müller cells. All above-mentioned questions could be answered with "yes." Specifically, the majority of Müller cells were labeled with antibodies directed to spm/spd, both within retinal sections and enzymatically isolated from retinal tissue. Müller cells in normal retinae express low levels of ODC but increase this expression markedly in cases of retinal pathology such as experimental epiretinal melanoma. Externally applied polyamines (1 mM) reduce (predominantly inward) whole-cell K+ currents, with the efficacies being spm > spd > put. If applied at the inside of membrane patches, spm (1 mM) blocks completely the outward currents through inwardly rectifying K+ channels but fails to affect the activity of large conductance, Ca2+-activated K+ channels. It is concluded that Müller cells contain endogenous channel-active polyamines, the synthesis of which may be up-regulated in pathological situations, and which may be involved in the control of both glial function and cell proliferation.

The Müller (glial) cell in normal and diseased retina: a case for single-cell electrophysiology.

In the retina of most vertebrates there exists only one type of macroglia, the Müller cell. Müller cells express voltage-gated ion channels, neurotransmitter receptors and various uptake carrier systems. These properties enable the Müller cells to control the activity of retinal neurons by regulating the extracellular concentration of neuroactive substances such as K+, GABA and glutamate. We show here how electrophysiological recordings from enzymatically dissociated mammalian Müller cells can be used to study these mechanisms. Müller cells from various species have Na(+)-dependent GABA uptake carriers, but only cells from primates have additional GABA receptors that activate Cl- channels. Application of glutamate analogues causes enhanced membrane currents recorded from Müller cells in situ but not from isolated cells. We show that mammalian Müller cells have no ionotropic glutamate receptors but respond to increased K+ release from glutamate-stimulated retinal neurons. This response is involved in extracellular K+ clearance and is mediated by voltage-gated (inwardly rectifying) K+ channels which are abundantly expressed by healthy Müller cells. In various cases of human retinal pathology, currents through these channels are strongly reduced or even extinguished. Another type of voltage-gated ion channels, observed in Müller cells from many mammalian species, are Na+ channels. In Müller cells from diseased human retinae, voltage-dependent Na+ currents were significantly increased in comparison to cells from control donors. Thus, the expression of glial ion channels seems to be controlled by neuronal signals. This interaction may be involved in the pathogenesis of retinal gliosis which inevitably accompanies any degeneration of retinal neurons. In particular, Müller cell proliferation may be triggered by mechanisms requiring the activation of Ca(2+)-dependent K+ channels. Ca(2+)-dependent K+ currents are easily elicitable in Müller cells from degenerating retinae and can be blocked by 1 mM TEA (tetraethylammonium). In purified Müller cell cultures, the application of 1 mM TEA greatly reduces the proliferative activity of the cells. These data clearly show that Müller cells are altered in cases of neuronal degeneration and may be crucially involved in pathogenetic mechanisms of the retina.

Effect of cutting the optic nerve on K+ currents in endfeet of Muller cells isolated from frog retina.

Membrane currents were recorded from Muller cells isolated from normal retinas and from retinas whose ganglion cell axons had been cut in the optic nerve 30-60 days previously. The surgical procedure did not block the retinal blood supply and did not allow the axons to regenerate. The principal finding was that after severing the optic nerve there was less inward rectification in response to voltage commands. That is, the maintained inward current (I K(IN)) produced in response to a hyperpolarizing voltage command decreased leading to a decrease in the ratio I K(IN)/I K(OUT) In 98 mM [K+]O, this ratio was 2.86 +/- 0.21 (mean +/- SE; n = 24) in controls and 1.13 +/- 0.13 (n = 21) in Muller cells from denervated retinae. Barium, a blocker of the potassium inward rectifier (I (KIR)), eliminated this difference. Moreover, severing the optic nerve also decreased the resting potentials of Muller cells in 2.5 mM [K+]O from -83 +/- 7 mV to -63 +/- 9 mV. The results suggest that the voltage-dependent behavior and selectivity of K+ inward rectifying channels (K (ir)) in the endfeet depends on the integrity of the closely apposed ganglion cells.

Potassium currents in endfeet of isolated Müller cells from the frog retina.

Voltage dependent potassium currents were recorded using the whole-cell mode of the patch-clamp technique for the first time from endfeet of Müller cells dissociated from the frog retina. Recordings from intact cells and isolated endfeet indicate that the inward rectifier potassium channel is the dominant ion channel in these cells and that the density of these channels is highest in the endfoot as has been previously reported for several other species. The present study uses rapid changes in [K+]o to understand the behavior of these channels in buffering [K+]o in the retina. With rapid changes in [K+]o, it was found that, at a membrane potential of -90mV, which is close to EK, increasing [K+]o from 3 to 10 mM produced an inward K+ current 5.48 +/- 0.89 SD (n = 9) times larger than outward current induced by decreasing [K+]o from 3 to 1 mM. The outward current was maximal at a holding potential of about -80mV and exhibited inactivation at more positive potentials. At -40 mV both the inward and outward currents are markedly reduced. The current voltage curve for the inward current was linear at holding potentials from -50 mV to -140 mV. Using 20 mV voltage steps, it was found that the voltage dependent K+ currents were unaffected by the addition of 2 mM Cd2+, a blocker of Ca(2+)-activated potassium currents, decreasing [Cl-]o from 120 mM to 5 mM or the substitution of 30 mM Na+ by TEA. The addition of 5 mM [Cs+]o blocked only the inward current. Both the outward and the inward currents disappeared in the absence of intracellular and extracellular K+; 0.3 mM [Ba2+]o blocked the inward current completely and strongly inhibited the outward current in a time and voltage dependent manner. We conclude that at physiological [K+]o and membrane potential, the K+ channels in the Müller cell endfoot are well suited to carry K+ both inward and outward across the membrane as required for spatial buffering.