Potassium Channels of Glial Cells - J-Stage

3 downloads 0 Views 698KB Size Report
+ buffering has been considered to couple tightly with water flux. Actually the aquaporin-4 water channel has been found to colocalize with Kir4.1 in retinal.
Jpn. J. Pharmacol. 87, 1 – 6 (2001)

REVIEW —Current Perspective— Potassium Channels of Glial Cells: Distribution and Function Yoshiyuki Horio* Department of Pharmacology, Sapporo Medical School, School of Medicine, South 1, West 17, Chu-ouku, Sapporo 060-8556, Japan Received January 26, 2001

ABSTRACT—Firing of neurons changes the extracellular concentration of K+ ions ([K+]o). Glial cells have the ability to maintain [K+]o at a constant level. This function has been called “K+-spatial buffering”. K+ channels are believed to be involved in K+-spatial buffering. Kir4.1 in retinal glial cells and Kir2.1, Kir2.3 and Kv1.5 in Schwann cells have been identified. All of these K+ channels show polarized distribution, which enables the channels to transport K+ ions to appropriate regions such as blood vessels and the vitreous body. These channels have a consensus C-terminal sequence that can bind a protein containing PDZ (PSD95/ dlg/ZO1) domains, which may regulate the distribution of the channels. Kir4.1 is predominantly expressed in membranes adjacent to basement membranes. Laminin, a component of basement membranes, is necessary for the surface expression of Kir4.1 in cultured retinal glial cells, suggesting that an extracellular signal regulates the function of glial cells. In some cases, K+ buffering has been considered to couple tightly with water flux. Actually the aquaporin-4 water channel has been found to colocalize with Kir4.1 in retinal glial cells. Recent studies of K+ channels have elucidated the mechanisms of old well-known phenomena and present new unknown roles of glial cells. Keywords: K+ channel, Glia, Water channel, PDZ domain, Laminin

Regulation of extracellular K+ concentration is indispensable for brain function Neural depolarization is terminated by the activity of K+ channels such as voltage-dependent and Ca2+-activated K+ channels, which are activated by increases of membrane potential and/or the intracellular Ca2+ concentration ([Ca2+]i). Because these channels transport intracellular K+ ions ([K+]i) to the extracellular space, the extracellular concentration of K+ ions ([K+]o) increases after neural excitation. The elevation of [K+]o has been demonstrated using K+-ion-selective microelectrodes and [42K+] flux. For example, the [K+]o at the retinal inner plexiform layer, which contains abundant synapses, increases prominently with exposure to a flash of light (1). Membrane potential of neurons is affected by [K+]o because neurons express inwardly rectifying K+ (Kir) channels such as Kir2.1, Kir2.2 and Kir2.3 (2, 3). When [K+]o increases, K+ ions are transported into the intracellular space through these Kir channels, resulting in elevation of the resting membrane potential. Too great an increase of [K+]o may induce uncontrolled hyperexcitability and abnormal synchronization of neurons.

Glial cells are indispensable for maintaining [K+]o Glial cells have been considered to regulate extracellular ionic homeostasis. To prevent unwanted neural excitation, glial cells aspirate excess extracellular K+ ions (Fig. 1). The pioneering work of Kuffler’s group demonstrated that glial cells possess high permeability to K+ ions, which enables these ions to intrude into and extrude from the cells. Moreover, they found that nerve impulses cause a slow depolarization of glial cells. Glial depolarization was observed from 50 – 150 ms after stimulation, followed by a decline that continued for seconds, while the nerve action currents lasted only 30 ms. They interpreted this to mean that glial cells aspirated excess K+ ions, which were liberated from neurons into the intercellular space, and were depolarized by the K+ ions (4). Based on this observation, they proposed the hypothesis that glia have a function to aspirate K+ ions at the sites where [K+]o is high, transport them within the cell and extrude the ions at the sites where [K+]o is low. They called this function of glia “K+-spatial buffering”. The retina has a simple and well-characterized structure of six layers. Because it can be isolated easily from the eyeball, the dynamics of K+ ion-movement can be characterized more easily than in the brain. Müller cells are the principal glial cells in the retina. In response

*Corresponding author. FAX: +81-11-612-5861 E-mail: [email protected] 1

2

Y. Horio

inhibited by barium, although a small part of the uptake is ouabain-sensitive (10). Extracellular cesium causes epileptiform activity characterized by synchronous burst discharges of pyramidal cells and prevents long-term depression of the CA1 hippocampal region in brain slices. Cesium enhances the increase in [K+]o evoked by stimulation. The target of cesium was found to be Kir channels of the astrocytes (11). Inhibition of K+ uptake through Kir channels of astrocytes induced the elevation of [K+]o. Although the contribution of the Na+-K+-2Cl– cotransporter to K+-uptake remains unclear, pharmacological tools suggest that K+ channels, especially Kir channels, dominantly participate in the intrusion of K+ ions.

Fig. 1. K+-spatial buffering of a glial cell. To regulate [K+]o, the glial cell aspirates K+ ions that are released from neurons, transport them within the cells and extrudes the ions at the site where [K+]o is low. K+ channels, which can be inhibited by barium and cesium, dominantly participate in the fluxes of K+ ions.

to photostimulation, a reduction in photoreceptor [K+]i is accompanied by an increase of [K+]i of Müller cells in the honeybee retina (5). Dissociated Müller cells have been shown to possess the capability to aspirate extracellular K+ ions from their distal ends and secrete cations from their proximal endfeet into the vitreous body (6). This phenomenon is named the K+-siphoning action. K+-uptake and regulation of [K+]o by glial cells were also demonstrated in situ using brain slices (7) (see below). Pharmacological experiments To resolve the mechanism of K+ uptake by glial cells, various inhibitors were used in cultured astrocytes and brain slices. High-K+ solution perfused via a microdialysis into the neocortex or hippocampus in brain slices causes only slight elevation of [K+]o at distances of 200 – 400 mm from the dialysis membrane. However, treatment with fluorocitric acid greatly raises [K+]o, suggesting that glial cells intrude K+ (7). Fluorocitric acid is one of gliotoxins and inhibits synthesis of isocitric acid from citiric acid in the TCA cycle. Rodent astrocytes can be dissociated and cultured easily. The accumulation of [K+]i in cultured astrocytes is inhibited by barium (8), ouabain and furosemide (9). Barium has been used as a tool to differentiate between passive and active uptake mechanisms and found mainly to block Kir channels. Ouabain and furosemide are inhibitors of Na+,K+-ATPase and the Na+-K+-2Cl– cotransporter, respectively. K+ uptake of glial cells induced by repetitive stimulation of the olfactory tract in brain slices is strongly

Patch-clamp experiments Among different ion channels, K+ channels comprise the most diverse group. To date, over 50 different K+ channels genes have been identified. Although studies using a patchclamp technique demonstrated expression of a voltagedependent K+ channel, a Ca2+-activated K+ channel and a Kir channel in glial cells, physiological properties of Kir channels also suggest that the channel is responsible for aspiration of K+ ions. Kir channels seem to be always active and they demand neither membrane depolarization nor an increase of [Ca2+]i. Moreover, the activity of Kir channels increases with the elevation of [K+]o. These properties enable glial cells to aspirate excess [K+]o ions. Rectification property of each Kir channel is different. Kir4.1, that has been identified in glia (see below), shows rather weaker rectification than that of strong inwardly rectifying K+ channels such as Kir2.1. Week rectification enables Kir4.1 to extrude K+ ions from the cell in some conditions. Abundant expression of Kir channels are found in endfeet of salamander and mammalian Müller cells. This polarized distribution of Kir channels and the evidence of K+-siphoning action suggest that some Kir channels also participate in extrusion of K+ ions (12 – 14). Kir4.1 In the retina, light closes cGMP-activated non-selective cation channels in the outer segment of photoreceptor cells and induces decrease of [K+]o in the distal retina. This decrease of [K+]o establishes a K+ efflux from the distal ends of Müller glial cells and a K+ influx into their proximal endfeet from the vitreous body. K+ ions also extrude from the apical regions of retinal pigment epithelial cells by light illumination. The K+ fluxes through Müller cells generate a slow PIII response in electroretinograms. Which Kir channel does participate in the mechanism? Fifteen different Kir channel genes have been isolated in mammals. Among them, only Kir4.1, which is also named Kir1.2 or KAB-2, was found in retinal Müller glial cells (13, 15) and retinal pigment epithelial cells (16). Immunoelectron microscopic

Glial K+ Channels

examination of Müller cells demonstrated the polarized expression of Kir4.1 on membranes adjacent to the vitreous body, pericytes and endothelial cells of capillaries (13, 14). This highly polarized distribution of Kir4.1 suggests that Kir4.1 participates in the net movement of K+ ions through membranes facing blood vessels and the vitreous body. Kir4.1 is also expressed in basal regions of microvilli of Müller cells, which face photoreceptor cells, and in the apical microvilli of retinal pigment epithelial cells (16). Recently, Kir4.1 knockout (Kir4.1 -/ -) mice were generated (17). Kir channel activity was not recorded in the Müller cells of Kir4.1 -/ - mice. The slow PIII wave in electroretinograms is totally absent in retinas from Kir4.1 -/- mice. Thus, the K+ flux of Müller cells induced by the change of [K+]o is mediated by Kir4.1. Expression of Kir4.1 is not limited to Müller cells. Satellite cells of cochlear ganglia, trigeminal ganglia and superior cervical ganglia also express Kir4.1 (18). Satellite glial cells wrap the somata of ganglion neurons with multiple layers of myelin sheaths, the structure of which is similar to that of myelin sheaths of Schwann cells. Developmental expression of Kir4.1 parallels the maturation of retinal and auditory activities. Kir4.1 is not detected at 1 to 5 days after birth. Weak immunoreactivity is found at 8 days, and it then increases during the following days, reaching the adult level at 14 days after birth (16, 18, 19). Although Kir4.1 is also expressed in some astrocytes of deep cerebellar nuclei and the hippocampus, Bergmann glia and oligodendrocytes of the cerebellum (20, 21), its expression is not detected in most astrocytes and oligodendrocytes or in satellite cells of the dorsal root and Auerbach’s ganglia (18, 21). In these cells, other types of Kir channels might be expressed in place of Kir4.1. Kir2.1 and Kir2.3 Schwann cells wrap themselves around axons to provide electrical insulation in the form of a myelin sheath in peripheral nerve fibers. In Schwann cells, K+ channels are not distributed homogeneously, but concentrated in the node of Ranvier (22). Kir2.1 and Kir2.3 were demonstrated to be localized at the microvilli of the nodes of Ranvier by using a specific antibody that recognizes both channels (23). The large surface area of microvilli in the nodal space suggests involvement with K+ buffering. Because Kir2.1 and Kir2.3 have strong inward rectification, these Kir channels may participate in the intrusion of K+ ions liberated from nerve axons. Kv1.5 Voltage-dependent K+ channels are responsible for generation of action potentials and for the control of neurotransmitter release in neurons and muscles. More than two dozen related genes have been cloned, all of these members are

3

activated in response to depolarization of a cell membrane. Glial cells have been considered to be electrically inexcitable cells. However, various types of glial cells express several voltage-dependent K+ channels. At present, roles of these K+ channels except Kv1.5 are unknown. Kv1.5, a member of Shaker subfamily, is expressed in Schwann cells. Kv1.5 is localized on the outer surface of Schwann cells in the vicinity of nodes of Ranvier (24). Kv1.5 channels are located at some distance from the microvilli of the nodes. Kv1.5 requires depolarization to be opened, and at depolarized potentials, K+ efflux would be favored. These results indicate that Schwann cells may aspirate K+ ions through Kir2.1 and Kir2.3 and release the ions from Kv1.5. Mechanisms for the localization of channels Kir4.1 has a polarized distribution on the membranes of Müller cells (13, 14). Immunoreactivity of Kir4.1 in isolated Müller cells is distributed in a clustered manner (13, 25). This clustered distribution may be important for Kir4.1 to be localized at the proper sites for intrusion and extrusion of K+ ions for the spacial buffering action. Then how do glial cells cluster Kir4.1 and localize the clusters of Kir4.1 in appropriate positions on their cell membranes? The primary amino acid sequence of Kir4.1 indicates that the carboxy-terminal region of this channel subunit ends with the sequence Ser-Asn-Val (20). Recently a protein family containing PDZ (PSD-95 / dlg/ ZO1) domains was shown to bind and cluster channels that have a sequence of Ser/ Thr-X-Val /Leu/Ile (X is any amino acid residue) in their C-terminal ends (26). Furthermore, in Drosophila, synaptic clustering of Shaker voltage-dependent K+ channels is abolished by mutations of dlg, a member of the PDZ protein family (27). Thus, the PDZ protein family might control cluster formation of Kir4.1. Actually, proteins containing PDZ domains could bind and cluster Kir4.1 (25). Because SAP97 is expressed in Müller cells and is distributed on the cell membranes in a clustered manner similar to Kir4.1, SAP97 is presumed to have a role in the distribution of Kir4.1. In addition to clustering, cotransfection of SAP97 causes a prominent increase of the Kir4.1 current magnitude, indicating that a protein containing PDZ domains might regulate the activity of the channel (25). In Schwann cells, expression of Kir2.1, Kir2.3 and Kv1.5 has been detected near the node of Ranvier. The C-terminal ends of these K+ channels possess a motif that can interact with PDZ proteins. However, at present, there is no information that a PDZ protein actually regulates the localization of these K+ channels in Schwann cells. The aquaporin-4 (AQP4) water channel is expressed in a polarized manner and colocalizes with Kir4.1 in retinal Müller cells (see below). AQP4 also has a C-terminal

4

Y. Horio

consensus sequence for the binding of PDZ domains; however, a PDZ protein that can interact with AQP4 has not been reported. The C-terminal ends of these glial channels are slightly different. Because each PDZ protein binds specific channels and receptors, different PDZ proteins may regulate specific localization of these channels in glial cells. Immunoreactivity of Kir4.1 is often found on the membranes in contact with the basement membranes. Does a component of basement membranes affect the localization of Kir4.1? Laminin is necessary for the cell-surface expression of Kir4.1 in Müller cells (13). Expression of Kir4.1 is not detected in the primary culture of Müller cells. Addition of insulin induces the expression of Kir4.1, but the channel is found in the cytosol. When Müller cells are cultured on laminin dishes in the presence of insulin, Kir4.1 is detected on the membranes. Thus, laminin has a pivotal role for the surface expression of Kir4.1 (Fig. 2). Polarized distribution of Kir4.1 in the Müller cells disappears when the cells are isolated. Immunohistochemical and patch clamp studies have shown diffuse distribution of Kir4.1 on the mem-

Fig. 2. Translocation of Kir4.1 into membranes by laminin. In the absence of laminin, Kir4.1 exists in the cytosol. Laminin, a component of basement membranes, is necessary for the surface expression of Kir4.1 in isolated Müller cells. The integrin family may bind laminin and transmit its signal, but the precise mechanism of translocation of Kir4.1 is still unknown.

branes of isolated Müller cells (15, 25). The diffusion of Kir4.1 may be due to the loss of the laminin signal, but the molecular mechanism of Kir4.1 translocation induced by laminin is still unknown. Volume regulation Extracellular volume changes dynamically during neuronal activity. Astrocyte cell volume transiently increases by exposure to high [K+]o solution (28). The K+ spatial buffering mechanism originally proposed by Orkand et al. (4) depends on the glial membrane selectively permeable to K+. Lack of anion permeability seems to normally occur in astrocytes, but colocarization of Cl– channels enables KCl uptake, which will then lead to water transport. The tight coupling among high neuronal activity, K+ buffering and water flux has been extensively studied in the brain and retina (for example, see ref. 29). Since K+ channels do not admit water, the water flux has to be mediated through a distinct channel. Glial cells express the water channel protein AQP4. AQP4 is expressed widely in the brain and retina, particularly at the brain-blood and brain-CSF interfaces. Immunoreactivities of AQP4 are distributed very similarly to those of Kir4.1 in retinal Müller glial cells. Double immunoelectron microscopic examination showed that AQP4 colocalized with Kir4.1 on the cell membranes adjacent to the vitreous body and blood vessels, suggesting that Kir4.1 and AQP4 cooperate in the siphoning of K+ ions and water in the retina (Fig. 3). At the microvilli of Müller cells, only Kir4.1 was expressed (14). Regulation of [K+]o can be achieved without volume change in these regions. Mice deficient in AQP4 (AQP4 -/-) had greatly reduced brain edema in response to acute water intoxication and ischemic stroke, suggesting that water is transported from blood vessels to astrocytes through the AQP4 water channel in these conditions (30). The expression of AQP4 in astrocytes was found in glial processes around the capillaries. Similarly, patch-clamp experiments showed that Kir channels are predominantly expressed at fine processes of astrocytes (31). If water is cotransported with K+ ions, Kir channels may participate in the promotion of brain edema. It is not known whether Kir4.1 -/ - mice show a phenotype similar to AQP4-deficient mice. Cl– ions have been considered to be cotransported with + K ions. Although several Cl– channels are expressed in glial cells, the Cl– channel that colocalizes with a K+ channel and a water channel has not been identified in glial cells. Conclusions and future perspectives Recent studies on glial channels have gradually revealed the molecular transport mechanisms of K+ ions and water. However, a number of questions still remain. Which kinds of K+ channels are expressed in the majority of astrocytes

Glial K+ Channels

5

Fig. 3. Colocalization of Kir4.1 and AQP4 on the membrane of the retinal glia. Double immunogold labeling of Kir4.1 (small particles) and AQP4 (large particles) shows the colocalization of both channels in the vitreal membrane of a Müller cell. The vitreal surface of A is enlarged in B. Asterisks indicate corresponding points. Note that only the vitreal side that faces the basement membrane expresses Kir4.1 and AQP4. Double arrow indicates vitral and the other aspects of the Müller cell endfoot. Small arrows indicate Kir4.1 labeling. M; Müller cell, Scale bar = A; 0.5 m m, B; 0.1 mm. Reproduced with permission from ref. 14 (Glia 26, 47 – 54, 1999, Copyright©1999, Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.)

and oligodendrocytes? Which kinds of Cl– channels colocalize with a K+ channel and a water channel? What is the physiological role for co-transport of water with K+ ions? Do glial K+ channels have some pathological role in brain edema and ischemia? Is there some regulation of channel activity? What is the role of extracellular matrix? Molecular techniques should contribute further to clarification of these and yet unknown roles of glial functions and lead to the development of novel drugs. Acknowledgments This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan.

6

7

8

9

REFERENCES 10 1 Oakley B II and Green DG: Correlation of light-induced changes in retinal extracellular potassium concentration with c-wave of the electroretinogram. J Neurophysiol 39, 1117 – 1133 (1976) 2 Horio Y, Morishige K, Takahashi N and Kurachi Y: Differential distribution of classical inwardly rectifying potassium channel mRNAs in the brain: comparison of IRK2 with IRK1 and IRK3. FEBS Lett 379, 239 – 243 (1996) 3 Karschin C, Dissmann E, Stuhmer W and Karschin A: IRK(1 – 3) and GIRK(1 – 4) inwardly rectifying K+ channel mRNAs are differentially expressed in the adult rat brain. J Neurosci 16, 3559 – 3570 (1996) 4 Orkand RK, Nicholls JG and Kuffler SW: Effect of nerve impulses on the membrane potential of glial cells in the central nervous system of amphibia. J Neurophysiol 29, 788 – 806 (1966) 5 Coles JA and Tsacopoulos M: Potassium activity in photoreceptors, glial cells and extracellular space of the drone retina:

11

12

13

14

Changes in photostimulation. J Physiol (Lond) 290, 525 – 549 (1979) Newman EA, Frambach DA and Odette LL: Control of extracellular potassium levels by retinal glial cell K+ siphoning. Science 225, 1174 – 1175 (1984) Largo C, Cuevas P, Somjen GG, Martin del Rio R and Herreras O: The effect of depressing glial function in rat brain in situ on ion homeostasis, synaptic transmission, and neuron survival. J Neurosci 16, 1219 – 1229 (1996) Walz W, Shargool M and Hertz L: Barium-induced inhibition of K+ transport mechanisms in cortical astrocytes – its possible contribution to the large Ba2+-evoked extracellular K+ signal in brain. Neuroscience 13, 945 – 949 (1984) Walz W and Hinks EC: Carrier-mediated KCl accumulation accompanied by water movements is involved in the control of physiological K+ levels by astrocytes. Brain Res 343, 44 – 51 (1985) Ballanyi K, Grafe P and Bruggencate GT: Ion activities and potassium uptake mechanisms of glial cells in guinea-pig olfactory cortex slices. J Physiol (Lond) 382, 159 – 174 (1987) Janigro D, Gasparini S, D’Ambrosio R, McKhann G II and DiFrancesco D: Reduction of K+ uptake in glia prevents longterm depression maintenance and causes epileptiform activity. J Neurosci 17, 2813 – 2824 (1997) Brew H, Gray PTA, Mobbs P and Attwell D: Endfeet of retinal glial cells have higher densities of ion channels that mediate K+ buffering. Nature 324, 466 – 468 (1986) Ishii M, Horio Y, Tada Y, Hibino H, Inanobe A, Ito M, Yamada M, Gotow T, Uchiyama Y and Kurachi Y: Expression and clustered distribution of an inwardly rectifying potassium channel, KAB-2 /Kir4.1, on mammalian retinal Müller cell membrane: their regulation by insulin and laminin signals. J Neurosci 17, 7725 – 7735 (1997) Nagelhus EA, Horio Y, Inanobe A, Fujita A, Haug FM, Nielsen S, Kurachi Y and Ottersen OP: Immunogold evidence suggests

6

15

16

17

18

19

20

21

22

Y. Horio that coupling of K+ siphoning and water transport in rat retinal Müller cells is mediated by a coenrichment of Kir4.1 and AQP4 in specific membrane domains. Glia 26, 47 – 54 (1999) Tada Y, Horio Y and Kurachi Y: Inwardly rectifying K+ channels in retinal Müller cells; comparison with the KAB-2 /Kir4.1 channel expressed in HEK293T cells. Jpn J Physiol 48, 71 – 80 (1998) Kusaka S and Puro DG: Intracellular ATP activates inwardly rectifying K+ channels in human and monkey retinal Müller (glial) cells. J Physiol (Lond) 500, 593 – 604 (1997) Kofuji P, Ceelen P, Zahs KR, Surbeck LW, Lester HA and Newman EA: Genetic inactivation of an inwardly rectifying potassium channel (Kir4.1 subunit) in mice: phenotypic impact in retina. J Neurosci 20, 5733 – 5740 (2000) Hibino H, Horio Y, Fujita A, Inanobe A, Doi K, Gotow T, Uchiyama Y, Kubo T and Kurachi Y: Expression of an inwardly rectifying K+ channel, Kir4.1, in satellite cells of rat cochlear ganglia. Am J Physiol 277, C638 – C644 (1999) Hibino H, Horio Y, Inanobe A, Doi K, Ito M, Yamada M, Gotow T, Uchiyama Y, Kawamura M, Kubo T and Kurachi Y: An ATP-dependent inwardly rectifying potassium channel, KAB-2 (Kir4.1), in cochlear stria vascularis of inner ear: its specific subcellular localization and correlation with the formation of endocochlear potential. J Neurosci 17, 4711 – 4721 (1997) Takumi T, Ishii T, Horio Y, Morishige K, Takahashi N, Yamada M, Yamashita T, Kiyama H, Sohmiya K, Nakanishi S and Kurachi Y: A novel ATP-dependent inward rectifier potassium channel expressed predominantly in glial cells. J Biol Chem 270, 16339 – 16346 (1995) Poopalasundaram S, Knott C, Shamotienko OG, Foran PG, Dolly JO, Ghiani CA, Gallo V and Wilkin GP: Glial heterogeneity in expression of the inwardly rectifying K+ channel, Kir4.1, in adult rat CNS. Glia 30, 362 – 372 (2000) Wilson GF and Chiu SY: Ion channels in axon and Schwann membranes at paranodes of mammalian myelinated fibers

studied with patch clamp. J Neurosci 10, 3263 – 3274 (1990) 23 Mi H, Deerinck TJ, Jones M, Ellisman MH and Schwarz TL: Inwardly rectifying K+ channels that may participate in K+ buffering are localized in microvilli (also called nodal processes) of Schwann cells. J Neurosci 16, 2421 – 2429 (1996) 24 Mi H, Deerinck TJ, Ellisman MH and Schwarz TL: Differential distribution of closely related potassium channels in rat Schwann cells. J Neurosci 15, 3761 – 3774 (1995) 25 Horio Y, Hibino H, Inanobe A, Yamada M, Ishii M, Tada Y, Satoh E, Hata Y, Takai Y and Kurachi Y: Clustering and enhanced activity of an inwardly rectifying potassium channel, Kir4.1, by an anchoring protein, PSD-95 /SAP90. J Biol Chem 272, 12885 – 12888 (1997) 26 Fujita A and Kurachi Y: SAP family proteins. Biochem Biophys Res Commun 269, 1 – 6 (2000) 27 Tejedor FJ, Bokhari A, Rojero O, Gorczyca M, Zhang J, Kim E, Sheng M and Budnik V: Essential role for dlg in synaptic clustering of Shaker K+ channels in vivo. J Neurosci 17, 152 – 159 (1997) 28 Walz W and Hinks EC: Carrier-mediated KCl accumulation accompanied by water movements is involved in the control of physiological K+ levels by astrocytes. Brain Res 343, 44 – 51 (1986) 29 Holthoff K and Witte OW: Intrinsic optical signals in rat neocortical slices measured with near-infrared dark-field microscopy reveal changes in extracellular space. J Neurosci 16, 2740 – 2749 (1996) 30 Manley GT, Fujimura M, Ma T, Noshita N, Filiz F, Bollen AW, Chan P and Verkman AS: Aquaporin-4 deletion in mice reduces brain edema after acute water intoxication and ischemic stroke. Nat Med 6, 159 – 163 (2000) 31 Barres BA, Koroshetz WJ, Chun LLY and Corey DP: Ion channel expression by white matter glia: the type-1 astrocyte. Neuron 5, 527 – 544 (1990)