Proliferative gliosis causes mislocation and inactivation of inwardly rectifying K+ (Kir) channels in rabbit retinal glial cells

Available online at www.sciencedirect.com

Experimental Eye Research 86 (2008) 305e313 www.elsevier.com/locate/yexer

Proliferative gliosis causes mislocation and inactivation of inwardly rectifying Kþ (Kir) channels in rabbit retinal glial cells Elke Ulbricht a, Thomas Pannicke a,*, Margrit Hollborn b,c, Maik Raap a, Iwona Goczalik a, Ianors Iandiev a,d, Wolfgang Ha¨rtig a, Susann Uhlmann d, Peter Wiedemann b, Andreas Reichenbach a, Andreas Bringmann b, Mike Francke a a

Paul Flechsig Institute of Brain Research, University of Leipzig Faculty of Medicine, Jahnallee 59, 04109 Leipzig, Germany b Department of Ophthalmology and Eye Clinic, University of Leipzig Faculty of Medicine, Leipzig, Germany c Interdisciplinary Center of Clinical Research (IZKF), University of Leipzig Faculty of Medicine, Leipzig, Germany d Translational Center for Regenerative Medicine, University of Leipzig, Leipzig, Germany Received 9 August 2007; accepted in revised form 2 November 2007 Available online 12 November 2007

Abstract Retinal glial (Mu¨ller) cells are proposed to mediate retinal potassium homeostasis predominantly by potassium transport through inwardly rectifying Kþ (Kir) channels. Retinal gliosis is often associated with a decrease in glial potassium conductance. To determine whether this decrease is caused by a downregulation of glial Kir channels, we investigated a rabbit model of proliferative vitreoretinopathy (PVR) which is known to be associated with proliferative gliosis. The membrane conductance of control Mu¨ller cells is characterized by large Kir currents whereas Mu¨ller cells of PVR retinas displayed an almost total absence of Kir currents. In control tissues, Kir2.1 immunoreactivity is localized in the inner stem processes and endfeet of Mu¨ller cells whereas Kir4.1 immunoreactivity is largely confined to the Mu¨ller cell endfeet. In PVR retinas, there is a mislocation of Kir channel proteins, with Kir4.1 immunoreactivity detectable in Mu¨ller cell fibers throughout the whole retina, and a decrease of immunoreactivity in the cellular endfeet. Real-time PCR analysis revealed no alteration of the Kir4.1 mRNA levels in PVR retinas as compared to the controls but a slight decrease in Kir2.1 mRNA. Western blotting showed no difference in the Kir4.1 protein content between control and PVR retinas. The data suggest that proliferative gliosis in the retina is associated with a functional inactivation of glial Kir channels that is not caused by a downregulation of the channel proteins but is associated with their mislocation in the cell membrane. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Kir4.1; Kir2.1; potassium current; gliosis; retina

1. Introduction

* Corresponding author. Tel.: þ49 341 972 5793; fax: þ49 341 972 5739. E-mail addresses: [email protected] (E. Ulbricht), [email protected] (T. Pannicke), margrit.hollborn@ medizin.uni-leipzig.de (M. Hollborn), [email protected] (M. Raap), [email protected] (I. Goczalik), yanors. [email protected] (I. Iandiev), [email protected] (W. Ha¨rtig), [email protected] (S. Uhlmann), peter.wiedemann@ uniklinik-leipzig.de (P. Wiedemann), [email protected] (A. Reichenbach), [email protected] (A. Bringmann), fram@medizin. uni-leipzig.de (M. Francke). 0014-4835/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.exer.2007.11.002

Glial cells play a major role in the ionic and osmotic homeostasis of the central nervous system. Neuronal activity causes local elevations in extracellular potassium which are buffered by glial cells. In the neural retina, Mu¨ller glial cells are suggested to take up excess potassium from the synaptic layers, and to release potassium ions into the blood and vitreous (Bringmann et al., 2006; Newman and Reichenbach, 1996). The transglial potassium transport is facilitated by inwardly rectifying potassium (Kir) channels expressed in the plasma membranes (Newman, 1993), and supported by a concomitant water flux through aquaporin water channels

306

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313

(Nagelhus et al., 1999). Among the various Kir channel subunits expressed by Mu¨ller cells (Raap et al., 2002), especially Kir4.1 and Kir2.1 channels have been implicated in mediating the potassium buffering currents (Kofuji et al., 2000, 2002). Both the potassium conductance and the Kir channels are unevenly distributed along the plasma membrane of Mu¨ller cells. The membrane conductance is high in subcellular domains facing extra-retinal fluid-filled spaces, i.e., the microvasculature and the vitreous body (Newman, 1987). The high potassium conductance of these cellular compartments has been explained by a high density of Kir4.1 channels localized in these membrane domains (Connors and Kofuji, 2002; Nagelhus et al., 1999). In contrast to Mu¨ller cells of vascularized retinas, Mu¨ller cells of avascular retinas such as of the rabbit and guinea pig, display the most prominent potassium conductance in the membranes of the cellular endfeet in the innermost retinal layers, facing the vitreous body (Newman, 1987). However, it is unknown whether this strongly polarized distribution of the potassium conductance in Mu¨ller cells of avascular retinas is also related to an asymmetric localization of Kir4.1 channels. Reactive gliosis in the retina is often associated with a decrease in the potassium conductance and a mislocation and/or downregulation of Kir4.1 channels in Mu¨ller cells (Francke et al., 2001; Pannicke et al., 2004, 2005, 2006). Until today, it is unclear whether the decrease in the potassium conductance of rabbit Mu¨ller cells is caused by a downregulation of Kir4.1. A downregulation or inactivation of Kir channels has been proposed to represent a hallmark of glial cellular dedifferentiation, and to be a precondition for Mu¨ller cell proliferation (Bringmann et al., 2000). Proliferative gliosis is associated with an almost complete absence of Kir currents in human Mu¨ller cells (Bringmann et al., 1999; Francke et al., 1997). Proliferative retinopathies (proliferative vitreoretinopathy, PVR) are a serious blinding complication of retinal detachment and can also be evoked by retinal surgery. PVR is characterized by the uncontrolled tumor-like proliferation of various cell types including glial cells, and results in the formation of periretinal contractile fibrocellular membranes that cause secondary tractional retinal detachment. Retinal detachment leads to photoreceptor degeneration and reactive changes in the inner retinal neurons (Fisher and Lewis, 2003). Here, we used a rabbit model of PVR to determine whether proliferative gliosis in the retina is associated with a mislocation and/or downregulation of Mu¨ller glial Kir channels. 2. Materials and methods 2.1. Animal model of PVR All experiments were done in accordance with the European Communities Council Directive 86/609/EEC, and were approved by the local authorities. Eight adult pigmented rabbits (2.5e3.5 kg; both sexes) were used. The animals were held under 12 h:12 h light/dark (day/night) room conditions, with free access to food and water. According to a method described previously (Francke et al., 2002), PVR was induced in

one eye of the animals while the other eye remained untreated and served as control. Anesthesia was induced by intramuscular ketamine (50 mg/kg) and xylazine (3 mg/kg; BayerVital, Leverkusen, Germany). The pupils of the eyes were dilated with topical tropicamide (1%; Ursapharm, Saarbru¨cken, Germany) and phenylephrine (5%; Ankerpharm, Rudolstadt, Germany). After pars plana sclerotomy, a circumscript vitrectomy was performed in the area of the future retinal detachment (in the ventro-nasal quadrant, below the medullary rays). A thin glass micropipette attached to a 250-ml Hamilton glass syringe was used to create a small local retinal detachment by injecting phosphate-buffered saline (pH 7.4) into the subretinal space. Another micropipette placed in the vitreous near the surface of the detached retina was used to inject 100 ml saline containing the proteolytic enzyme, dispase I (0.5 U; Boehringer, Mannheim, Germany). After injections, the sclerotomies and the conjunctiva were closed. Two weeks after operation, the animals were anaesthetized as above, and killed by intravenous T61 (3 ml; embutramid 0.2 g/ml, mebezonium iodide 0.05 g/ml, tetracain hydrochloride 5 mg/ml; Hoechst, Unterschleißheim, Germany), and the eyes were removed. 2.2. Electrophysiological recordings Pieces of retinal tissue were incubated in papain (0.2 mg/ml; Roche Molecular Biochemicals)-containing calcium- and magnesium-free phosphate-buffered saline (pH 7.4) for 30 min at 37  C, followed by several washing steps with normal saline. After short incubation in saline supplemented with DNase I (200 U/ml; SigmaeAldrich, Taufkirchen, Germany), the tissue pieces were triturated by a wide-pore pipette, to obtain suspensions of isolated cells. The cells were stored at 4  C in serum-free minimum essential medium until use within 4 h after cell isolation. Membrane currents of acutely isolated Mu¨ller cells were recorded in the whole-cell configuration of the patch-clamp technique. Voltage-clamp records were performed at room temperature (22  C) using the Axopatch 200A amplifier (Axon Instruments, Foster City, CA, USA) and the ISO-2 software (MFK, Niedernhausen, Germany). The signals were lowpass filtered at 1 kHz (eight-pole Bessel filter) and digitized at 5 kHz, using a 12-bit A/D converter. Patch pipettes were pulled from borosilicate glass (Science Products, Hofheim, Germany) and had resistances between 4 and 6 MU when filled with the intracellular solution containing (in mM) 10 NaCl, 130 KCl, 1 CaCl2, 2 MgCl2, 10 ethyleneglycolbis(aminoethylether)tetra-acetate, and 10 N-2-hydroxyethyl-piperazine-N0 -2-ethanesulfonic acid (HEPES) adjusted to pH 7.1 with Tris-base. The recording chamber was continuously perfused with extracellular solution which contained (in mM) 135 NaCl, 3 KCl, 2 CaCl2, 1 MgCl2, 1 Na2HPO4, 10 HEPES-Tris, and 11 glucose (pH 7.4). To evoke potassium currents, depolarizing and hyperpolarizing voltage steps of 250 ms duration, with increments of 10 mV, were applied from a holding potential of 80 mV. The membrane capacitance of the cells was measured by the integral of the uncompensated capacitive artifact (filtered at 6 kHz) evoked by a hyperpolarizing voltage

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313

step from 80 to 90 mV when Ba2þ (1 mM) was present in the bath solution. The steady-state whole-cell currents were measured at the end of 250-ms voltage steps, and the amplitude of Kir currents was estimated as the difference between amplitudes recorded at the voltage steps to 100 mV and to 160 mV. Statistical analysis (unpaired Student’s t-test) was performed using the SigmaPlot program (Jandel Corp.). 2.3. Immunohistochemistry Isolated retinas were fixed in 4% paraformaldehyde for 30 min. The tissues were embedded in 3% agarose, and sections of 80 mm thickness were made with a vibratome. After permeabilization with 1% Triton X-100/ 3% dimethylsulfoxide (DMSO) in saline at 37  C for 30 min and blocking the unspecific binding sites with normal donkey serum (5%; Dianova, Hamburg, Germany), the slices were incubated with the primary antibodies at 4  C for 12 h. After washing in Triton/DMSO in saline, the secondary antibody was applied for 1 h at room temperature. Cell nuclei were stained with Hoechst 33258 (Invitrogen, Paisley, UK). Unspecific staining of secondary antibodies was verified by omitting primary antibodies. The following antibodies were used: goat antiKir4.1 (1:200; clone G-19; Santa Cruz Biotechnology, Santa Cruz, CA, USA), goat anti-Kir2.1 (1:200; clone C-20; Santa Cruz), mouse anti-glial fibrillary acidic protein (GFAP)-Cy3 (1:400; Ga5 clone; SigmaeAldrich), and Cy2-conjugated donkey anti-goat IgG (1:1000, Dianova). Images were taken with a confocal laser scanning microscope (LSM 510 Meta; Zeiss, Oberkochen, Germany). 2.4. Western blotting Isolated retinal pieces were homogenized in lysis buffer (0.5 M EDTA in aqua dest. with Tris base, pH 8.0) with protease inhibitors, subsequently centrifuged and rehomogenized in RIPA buffer. The homogenates were centrifuged once more at 13,000  g for 45 min at 4  C. The protein concentration was determined with the method of Bradford. Equal amounts of protein (10 mg) were separated on 12% SDSepolyacrylamide gel. After electrophoresis, the proteins were transferred to nitrocellulose membranes and immunoblotted with a rabbit anti-Kir 4.1 antibody (1:500; SigmaeAldrich) for 14 h at 4  C. The membranes were washed, incubated with alkaline phosphatase-conjugated anti-rabbit IgG (1:5000; Sigmae Aldrich) for 1 h at room temperature, and visualized with alkaline phosphatase development tablets (SigmaeAldrich). 2.5. Real-time PCR and sequence analysis Total RNA was extracted from retinal pieces using Trizol reagent (Invitrogen) and purified with the RNeasy Mini Kit (Qiagen, Hilden, Germany) and DNase I (Roche, Mannheim, Germany). The quality of the RNA was analyzed by agarose gel electrophoresis. cDNA was synthesized with 1 mg of total RNA using the RevertAid H Minus First Strand cDNA Synthesis Kit (Fermentas, St. Leon-Roth, Germany).

307

Semi-quantitative real-time RTePCR was performed with the MyIQ-Single-Color Real-Time PCR Detection System (BioRad, Munich, Germany). The primer pairs were selected according to the published rabbit cDNA sequences and homologue sequences of other mammalians. The following primer pairs were used: glyceraldehyde 3-phosphate dehydrogenase (GAPDH; AB231852/rabbit), forward 50 -AGGTCATCCAC GACCACTTC-30 , reverse 50 -GTGAGTTTCCCGTTCAGC TC-30 ; Kir4.1 (AB005734/rabbit), forward 50 -GGGACCTG CTGGAGTTGGGAC-30 , reverse 50 -TCACCACTGCGAA GAGGGAGG-30 ; Kir2.1 (AF021138/rabbit), forward 50 GAGCACAGCTCCTCAAATCC-30 , reverse 50 -TGCATTGT CCATGTCCTGTT-30 ; Kir7.1 (AJ006128/human, AF200713/ guinea pig, AJ006129/rat), forward 50 -GTCCAAGTGCA ATCGCCTTA-30 , reverse 50 -CGGTAGGTAGGATGTCC TCC-30 . The PCR solution contained 1 ml cDNA, specific primer set (1 mM each), and 10 ml of QuantiTect SYBR Green PCR Kit (Qiagen) in a final volume of 20 ml. The PCR parameters were initial denaturation and enzyme activation (one cycle at 95  C for 15 min); denaturation, annealing, amplification and quantification, 45 cycles at 95  C for 30 s, 60  C for 30 s, and 72  C for 60 s; melting curve, 55  C with the temperature gradually increased (0.5 K) up to 95  C. For Kir7.1, the amplification time was 90 s. The mRNA expression was normalized to the levels of GAPDH mRNA, and the changes were calculated as described (Pfaffl, 2001). The real-time PCR efficiency (E) was calculated according to the equation: E ¼ 10[1/slope] (Rasmussen, 2001). The efficiencies of the target genes were found to be similar in the investigated range between 0.5 and 100 ng cDNA input (GAPDH, 2.12; Kir2.1, 2.12; Kir4.1, 2.16; Kir7.1, 1.99). Since the sequence of rabbit Kir7.1 is not available, the PCR fragment was subcloned into the pT-Adv cloning vector (Clontech, Heidelberg, Germany) and sequenced. Achieved sequence information was submitted to the NCBI gene bank (accession no. AJ309933). We performed also a sequence analysis of the carboxy terminus of rabbit Kir4.1. The partial sequence information was used to perform RACEePCR to identify the 50 and 30 ends. The PCR products were ligated into the pT-Adv cloning vector, transformed into Escherichia coli DH5, and sequenced. 3. Results 3.1. Morphological / histological examination To evoke proliferative gliosis in the rabbit retina, we induced a small circumscribed retinal detachment by injection of saline into the subretinal space and subsequently applied a proteolytic enzyme into the vitreous near the detached retina (Francke et al., 2002). Within 2 weeks after surgery, a massive PVR developed, mainly characterized by the formation of epiretinal fibrocellular membranes (Fig. 1). One major constituent of such membranes are retinal glial cells that proliferate and migrate onto both surfaces of the neural retina (Nork et al., 1987). The contraction of the cellular membranes leads to further retinal detachment (Fig. 1). The detachment of the neural

308

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313

Fig. 1. Morphological alterations of the retina in PVR eyes. The images show open eyes (cornea and lens were removed). (A) Control eye. Rabbit retinas are avascular, with the exception of the medullary rays (mr). (B) PVR eye. PVR caused the development of fibrocellular epiretinal membranes (arrowhead), secondary retinal detachment (arrow), and a partial disruption of the blood vessels in the medullary rays. The retinal tear with wrinkled edges (asterisk) developed from the surgically induced retinal detachment.

retina from the pigment epithelium caused a full degeneration of photoreceptor segments and the death of most photoreceptor cells, as indicated by the decrease in the number of cell nuclei in the outer nuclear layer of Hoechst-stained slices of PVR retinas compared to control (Fig. 2A). In addition, the PVR was associated with degenerative changes in the inner retinal tissue, as indicated by the disorganization of the inner nuclear layer and the decrease in the thickness of the inner plexiform (synaptic) layer (Fig. 2A). The activation of Mu¨ller glial cells is indicated by the strong upregulation of the glial intermediate filament protein, glial fibrillary acidic protein (GFAP; Fig. 2B). Control tissues were not labeled with the antiGFAP antibody, whereas Mu¨ller cell fibers traversing the whole retinal thickness displayed a strong GFAP labeling in PVR retinas. It is known that GFAP is expressed by retinal astrocytes in mammalian retinas, whereas Mu¨ller cells normally express GFAP only to a marginal extent but upregulate GFAP under pathological conditions (Bringmann et al., 2006). The absence of astrocytes in the avascular retina of the rabbit (with the exception of the medullary rays: Fig. 1) explains the lack of GFAP labeling in control tissues (cf. Schnitzer, 1985). 3.2. Potassium currents of Mu¨ller glial cells Proliferative gliosis of Mu¨ller cells is associated with a reduction in the Kir channel-mediated potassium conductance (Bringmann et al., 1999; Francke et al., 1997). We compared the potassium conductance of Mu¨ller cells from control and PVR retinas using whole-cell patch-clamp recordings of acutely isolated cells. Examples of such recordings are shown in Fig. 3A. Mu¨ller cells of control retinas displayed large inward and outward potassium currents when the membrane potential was stepped to hyper- and depolarizing potentials from a holding potential of 80 mV, near the resting membrane potential. The currentevoltage relation of the potassium

conductance displayed a weak inward rectification around the resting membrane potential (Fig. 3B), reflecting the importance of weakly rectifying Kir4.1 channels as contributors to the whole-cell potassium conductance. Application of the Kir channel blocker, barium chloride, to the bath solution resulted in a disappearance of the inward currents (Fig. 3A) while outwardly rectifying potassium currents which activated at potentials positive to 40 mV (Fig. 3B) remained present. Mu¨ller cells of the PVR retina displayed a virtually complete absence of inward currents (Fig. 3A,B). In the mean, the density of the inward currents decreased to 5.7  4.7% of control (100%; P < 0.001; Fig. 3C). In contrast, the membrane capacitance increased significantly in Mu¨ller cells from PVR retinas (85  33 pF, n ¼ 29) compared to control cells (55  20 pF, n ¼ 25; P < 0.001) indicating hypertrophy of gliotic Mu¨ller cells. The reduction of inward potassium currents was associated with a decrease of the resting membrane potential, in the mean from 81.0  6.2 mV in control cells to 62.5  16.5 mV in cells from PVR retinas (P < 0.001). The data indicate that proliferative gliosis of Mu¨ller cells is associated with an alteration in the profile of the membrane conductance, from a current pattern with prominent Kir currents to a pattern with dominant voltage-dependent, outwardly rectifying potassium currents. 3.3. Retinal distribution of Kir channel proteins We stained retinal slices to determine the distribution of two Kir channel subunits in the rabbit retina. Immunoreactivity for Kir2.1 was found throughout all retinal layers (Fig. 2C). In the inner retina, Kir2.1 was localized to the stem processes of Mu¨ller cells traversing the inner plexiform layer, and was enriched in the Mu¨ller cell endfeet in the nerve fiber and ganglion cell layers. In the outer retina, Kir2.1 immunoreactivity was found in the outer plexiform layer and in photoreceptor

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313

309

Fig. 2. Proliferative gliosis in the rabbit retina is associated with a mislocation of Kir channels. Slices of control (above) and PVR retinas (below) were immunostained against GFAP (B), Kir2.1 (C), and Kir4.1 (D), respectively. The inserts in C and D display Kir2.1 and Kir4.1 immunoreactivity, respectively, in the ganglion cell layer from other retinal areas at higher magnification. Labeled Mu¨ller cell endfeet envelop immunonegative ganglion cell somata. Cell nuclei were labeled with Hoechst 33258 (A). (E) Specificity controls with omitted primary antibodies. GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; ONL, outer nuclear layer; OPL, outer plexiform layer; PRS, photoreceptor segments. Note the absence of photoreceptor segments in PVR retinas due to retinal detachment (C). The bar (20 mm) is valid for all images except the inserts.

segments. Apparently, Kir2.1 is expressed by Mu¨ller cells in the inner retina and by photoreceptor cells in the outer retina. An expression of Kir2.1 by Mu¨ller cells has been already described in the murine retina (Kofuji et al., 2002). In PVR retinas, Kir2.1 immunoreactivity remained localized in Mu¨ller cell fibers throughout all retinal layers (Fig. 2C). There was an apparent decrease of the Kir2.1 labeling intensity in the endfeet of Mu¨ller cells. The disappearance of Kir2.1 label in the outer plexiform layer and photoreceptor segments reflects the degeneration of photoreceptor cells due to retinal detachment, as a consequence of PVR. Although the expression of Kir2.1 in photoreceptor cells is suggested by Kir2.1 immunoreactivity, detailed investigations including tests of physiological function of the Kir2.1 channel in photoreceptor cells are beyond the scope of this study, and thus the question of Kir2.1 channels in photoreceptor cells remains to be elucidated. Immunoreactivity for Kir4.1 was found in the inner part of the retina where it was strongly enriched in the Mu¨ller cell endfeet (Fig. 2D). This polarized distribution of Kir4.1 is in good agreement with the previously described distribution of the potassium conductance of rabbit Mu¨ller cells which shows the greatest amplitude in the membranes of the endfeet (Newman, 1987; Reichenbach and Eberhardt, 1986). In PVR

retinas, there was a conspicuous alteration in the retinal distribution of Kir4.1 (Fig. 2D). Here, the immunoreactivity for Kir4.1 was localized to the whole Mu¨ller cell fibers traversing the retinal tissue from the inner to the outer limiting membranes. The redistribution of Kir4.1 was associated with a decrease in the Kir4.1 staining of the Mu¨ller cell endfeet. Double labeling revealed that the immunoreactivities for Kir2.1 and Kir4.1 were localized to GFAP-expressing Mu¨ller cell fibers (not shown). The data suggest that proliferative gliosis in the retina is associated with an alteration in the location of glial Kir channels; especially the Kir4.1 protein displays a redistribution from prominent expression in Mu¨ller cell endfeet to random distribution along the entire length of Mu¨ller cells. 3.4. Gene and protein expression of Kir channels To determine whether the decrease in the potassium conductance of Mu¨ller cells from PVR retinas (Fig. 3) was caused by a downregulation of glial Kir channel subunits, we performed real-time PCR and immunoblot analyses. It has been shown that the neural retina of the guinea pig expresses a diversity of Kir channel subunits including Kir2.1, Kir4.1, and Kir7.1 (Raap et al., 2002). We found an expression of gene

310

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313

Fig. 3. Proliferative gliosis in the rabbit retina is associated with a strong reduction in glial inward potassium conductance. (A) Representative examples of current traces that were recorded in isolated Mu¨ller cells of a control retina (in the absence and presence of 1 mM barium chloride) and a PVR retina. Note the absence of inward currents (downwardly depicted) in the presence of barium and in the cell of the PVR retina. Whole-cell currents were evoked by 20-mV incremental voltage steps between 160 and þ20 mV, from a holding potential of 80 mV. (B) Currentevoltage relationships of the cells from control and PVR retinas. Current amplitudes were measured at the end of 250-ms voltage steps. (C) Mean (S.D.) density of the inward currents of Mu¨ller cells isolated from control (n ¼ 57) and PVR retinas (n ¼ 29). The currents were measured between the voltage steps from 80 to 100 and to 160 mV, respectively. Significant difference vs. control: *P < 0.001.

transcripts of these Kir channel subunits in the rabbit retina (Fig. 4A). In control retinas, the expression level of Kir4.1 was 11.9-fold higher than the level of Kir2.1, and the mRNA level of Kir7.1 was 12.9-fold higher compared to Kir2.1. We found that the gene expression levels of Kir4.1 were not different between PVR and control retinas (Fig. 4B). The expression level of Kir2.1 was slightly reduced by 27.7  19.3% in PVR retinas as compared to the controls (P < 0.05). A strong downregulation (by 77.1  7.3%; P < 0.01) was found for Kir7.1 in PVR retinas (Fig. 4B). In Western blots of retina lysates, the anti-Kir4.1 antibody stained a prominent band at w42 kDa (Fig. 4C), likely reflecting the monomeric form of Kir4.1. There was no difference in the Kir4.1 protein content between control and PVR retinas (Fig. 4C). The data suggest that proliferative gliosis of rabbit Mu¨ller cells is not associated with alterations in the transcript and protein levels of Kir4.1. 3.5. Kir channel sequences The anti-Kir4.1 antibodies used in the present study were raised against the carboxy terminal peptide of rat Kir4.1 which

is homologue to human and mouse peptides. To prove whether these antibodies are feasible to stain rabbit tissues, we sequenced the carboxy terminus of rabbit Kir4.1. We found that the amino acid sequence of rabbit Kir4.1 differed in only one amino acid from human Kir4.1 (not shown), suggesting that the antibodies can be used in rabbit tissues. Furthermore, we partially sequenced rabbit Kir7.1. When compared to human Kir7.1 (Altschul et al., 1997), the rabbit homologue showed 90% identity in the nucleotide sequence, and 93% identity in the amino acid sequence (not shown). 4. Discussion Proliferative gliosis in the retina is associated with a de-differentiation and proliferation of glial cells. A major differentiation marker of retinal glial cells is the expression level of Kir channels (Bringmann et al., 2000) which are implicated in the spatial buffering of extracellular potassium (Bringmann et al., 2006; Newman and Reichenbach, 1996). During proliferative gliosis, the Kir channel-mediated potassium conductance of retinal glial cells is almost completely absent. The mechanism

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313

311

Fig. 4. The decrease in glial potassium conductance is not caused by downregulation of Kir4.1. (A) RTePCR was carried out to determine the presence of mRNAs for Kir2.1, Kir4.1, Kir7.1, and GAPDH in tissues of a control (co) and a PVR retina. The negative control () was done by adding water instead of cDNA. The presence of gene products was determined after 45 cycles. (B) Mean (S.E.M.) relative mRNA levels of Kir channels in PVR retinas (n ¼ 5) as compared to control retinas (n ¼ 5). The data were obtained with real-time PCR. (C) Western blots of Kir4.1 protein in tissues of a control (co) and a PVR retina. The blots are representative for three independent experiments. Significant differences vs. control: *P < 0.05; **P < 0.01.

of this alteration is unclear. The main Kir channel subunit of retinal glial cells is Kir4.1 (Kofuji et al., 2000) which, in the rabbit retina, is expressed at a w12-fold higher level than Kir2.1. We show that the decrease in the retinal glial potassium conductance is not caused by a downregulation of Kir4.1, but is likely related to a mislocation of channel proteins which may lead to a functional inactivation of the channels. In control retinas, Kir2.1 channels are localized in membrane domains of Mu¨ller cells that abut retinal neurons, through which the Mu¨ller cells take up excess potassium released from active neurons (both synaptic layers and ganglion cell layer). On the other hand, the localization of Kir4.1 channels is largely restricted to the membranes of the Mu¨ller cell endfeet in the nerve fiber and ganglion cell layers. In gliotic Mu¨ller cells, the distribution of Kir4.1 protein is altered into an even localization along the whole Mu¨ller cell which is associated with a decrease in the prominent expression of Kir4.1 in Mu¨ller cell endfeet. We found that the gene and protein expression of Kir4.1 was not different between control and PVR retinas, suggesting that the mislocation is related to an alteration in channel clustering and/or in the association of the Kir4.1 subunit with putative binding partners such as aquaporin water channels, other Kir subunits, PDZ domaincontaining syntrophins, and laminin (Connors and Kofuji, 2002, 2006; Dalloz et al., 2003; Ishii et al., 2003; Noel et al., 2005). Nevertheless, the work of Connors and Kofuji (2002) and of Dalloz et al. (2003) showed that the mislocation of Kir4.1 in mouse retinas did not inevitably result in the functional inactivation of this channel. On the other hand, it has been reported by Pannicke et al. (2004, 2005, 2006) that in rat retinas three different pathological models of reactive gliosis (ischemia, ocular inflammation and diabetes) could induce a functional inactivation of the Kir4.1 channel that was accompanied with the downregulation of the Kir4.1 channel protein expression as judged by immunofluorescence or Western blot studies. These differences in the effects of gliosis on Kir4.1 channels

may reflect species differences of rabbit retinas compared to rat or mouse retinas, respectively. Rats and mice possess vascular retinas, in contrast to the avascular retina of rabbits (with exception of a central area in the medullary rays), i.e., rabbit Mu¨ller cells do not contact blood vessels, and release potassium mainly via their endfeet into the vitreous. A breakdown of the blood-retinal barrier, with a subsequent diffusion of various factors into the retinal tissue will not occur in the rabbit retina (but in the rat retina). Kir channel function and interactions with associated proteins of the Mu¨ller cell cytoskeleton might be different depending on retinal vascularization. Mu¨ller cells undergo reactive gliosis in various cases of retinal injury or disease (see Bringmann et al., 2006 for a recent review). However, the consequences may range from increases in the expression of intermediate filaments and cell hypertrophy to alterations in physiological functions and massive proliferative gliosis. At present it is unknown how the variability of the glial reaction is determined; the progression rate of the injury (Iandiev et al., 2006) was assumed to play a role. Although proliferation may occur in ischemic injury or diabetic retinopathy, it is particularly prominent in PVR, suggesting a correlation of Kir4.1 channel inactivation and proliferation of Mu¨ller cells (Bringmann et al., 2000). Further studies are necessary to investigate the diversity of Mu¨ller cell reactions in different cases of retinal degeneration. However, our data suggest that clustering of the channels or their association with putative binding partners might regulate the activation state of Kir4.1 channels, resulting in full inactivation when the association or clustering is lost. Based on the present data, it can not unequivocally be decided whether the Kir4.1 subunits are correctly inserted into the cell membrane. It may be possible that the immunoreactivity derives from channel subunits located in the cytoplasm of the cells. The mechanisms of the mislocation of Kir channels during proliferative gliosis are not yet clear. In cultured Mu¨ller cells, the presence of the basal lamina constituent, laminin and of the

312

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313

PDZ-ligand domain of Kir4.1 were found to be necessary to induce plasma membrane clustering of Kir4.1 (Ishii et al., 1997; Noel et al., 2005). The membrane localization of Kir4.1 is stabilized by Dp71 (Connors and Kofuji, 2002; Dalloz et al., 2003), but is not regulated by the PDZ domain protein a-syntrophin (Puwarawuttipanit et al., 2006) whereas an involvement of b-syntrophins in the membrane anchoring of Kir4.1 cannot be ruled out. It remains to be determined whether the dispersion of Kir4.1 channel clusters is sufficient to inactivate the channels in rabbit Mu¨ller cells, or whether additional signals, such as alterations in the actin cytoskeleton or in the channel phosphorylation state, must be present to cause functional inactivation of the channels. Moreover, it can not be ruled out that a decrease in the expression of Kir2.1 contributes to the reduction in glial potassium conductance. However, since the decrease in Kir2.1 mRNA level was only w30% whereas the inward potassium currents were virtually absent in Mu¨ller cells from PVR retinas, we assume that a decrease in the expression of Kir2.1 contributes only marginally to the decrease of the potassium conductance. This assumption is supported by the previous observation that Kir4.1 channels represent the major conductive pathway contributing to the potassium conductance of isolated Mu¨ller cells (Kofuji et al., 2000); the importance of Kir4.1mediated currents is also reflected in the weak rectification of the current-voltage relation of Mu¨ller cell currents at the resting membrane potential of w80 mV (Fig. 2B). In this study we observed a strong downregulation of Kir7.1 in PVR retinas as compared to the controls. However, it is unclear which retinal cell types express Kir7.1. In the bovine retina, Kir7.1 immunoreactivity has been localized to the pigment epithelium and to cell bodies in the inner nuclear layer (Yang et al., 2003), likely reflecting an expression of Kir7.1 by retinal second-order neurons. The assumption of a neuronal (and not a glial) expression of Kir7.1 is supported by the previous finding that Kir7.1 mRNA is present in the retinal tissue but not in primarily cultured Mu¨ller cells of the guinea pig (while transcripts for Kir2.1 and Kir4.1 were present in both the retinal tissue and cultured Mu¨ller cells) (Raap et al., 2002). Thus, the downregulation of Kir7.1 likely reflects the degeneration of retinal second order neurons which is also indicated by the disaggregation of the inner nuclear layer and by the reduced thickness of the inner plexiform layer in PVR retinas. The loss of glial inward potassium conductance should disturb the potassium clearance from the diseased retinal tissue. An extracellular accumulation of potassium will result in neuronal hyperexcitation and glutamate toxicity. Excitotoxicity has been suggested to be one factor underlying the morphologic and biochemical alterations in the retina after detachment (Fisher and Lewis, 2003; Marc et al., 1998), and may explain (at least in part) the structural alterations of the retina found in the present study. In retinas of Dp71-null mice, a mislocation of Kir4.1 in Mu¨ller cells has been found to be associated with a greater vulnerability of retinal ganglion cells to ischemia-reperfusion injury (Dalloz et al., 2003). Since excitotoxicity is a major factor involved in ganglion cell death after

ischemia (Osborne et al., 2004), a mislocation and inactivation of glial Kir channels may have an important pathogenic impact contributing to excitotoxic retinal degeneration under various different pathological conditions. Kir2.1 channels are strongly rectifying channels that mediate predominantly inward currents and almost no outward currents (Kubo et al., 1993) whereas Kir4.1 channels are weakly rectifying, which means that they mediate outward and inward currents with similar amplitudes (Takumi et al., 1995). Therefore, the main pathway of potassium extrusion from Mu¨ller cells are Kir4.1 channels which are selectively located in the membranes of the Mu¨ller cell endfeet in the rabbit retina. It is known that Kir4.1 and aquaporin-4 water channels are colocalized in membrane domains of Mu¨ller cells (Nagelhus et al., 1999); this finding has led to the suggestion that the potassium transport through Mu¨ller cells is coupled to a water transport that mediates fast activity-dependent water clearance from the retinal tissue. Thus, inactivation of Kir4.1 channels should disturb both the potassium and the water transport through Mu¨ller cells; the resulting water accumulation within the retinal tissue will contribute to the development of tissue edema. Indeed, in a previous study we found that experimental detachment of the rabbit retina causes the formation of fluidfilled cystoid spaces selectively in the nerve fiber and ganglion cell layers (Faude et al., 2001). It has been suggested that the formation of these edematous spaces is the result of the decrease in Kir4.1 in Mu¨ller cell endfeet which should disturb the water clearance from the innermost retinal layers (Francke et al., 2005). Moreover, an impairment of the potassium and water transport through Mu¨ller cells has been causally linked to the degeneration of photoreceptor cells (Francke et al., 2005). Further research is necessary to reveal the pathogenic role of Kir channel mislocation in retinal degeneration and induction of glial cell proliferation.

Acknowledgments This work was supported by grants from the: German Federal Ministry of Education and Research BMBF (Grant number: 1113MB); Interdisciplinary Center of Clinical Research (IZKF) at the University of Leipzig Faculty of Medicine (Grant numbers: C5, C35); and Deutsche Forschungsgemeinschaft (Grant numbers: GRK 1097, RE 849/8, KO 1547/4-1).

References Altschul, S.F., Madden, T.L., Scha¨ffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389e3402. Bringmann, A., Francke, M., Pannicke, T., Biedermann, B., Faude, F., Enzmann, V., Wiedemann, P., Reichelt, W., Reichenbach, A., 1999. Human Mu¨ller glial cells: altered potassium channel activity in proliferative vitreoretinopathy. Invest. Ophthalmol. Vis. Sci. 40, 3316e3323. Bringmann, A., Francke, M., Pannicke, T., Biedermann, B., Kodal, H., Faude, F., Reichelt, W., Reichenbach, A., 2000. Role of glial Kþ channels in ontogeny and gliosis: A hypothesis based upon studies on Mu¨ller cells. Glia 29, 35e44.

E. Ulbricht et al. / Experimental Eye Research 86 (2008) 305e313 Bringmann, A., Pannicke, T., Grosche, J., Francke, M., Wiedemann, P., Skatchkov, S.N., Osborne, N.N., Reichenbach, A., 2006. Mu¨ller cells in the healthy and diseased retina. Prog. Retin. Eye Res. 25, 397e424. Connors, N.C., Kofuji, P., 2002. Dystrophin Dp71 is critical for the clustered localization of potassium channels in retinal glial cells. J. Neurosci 22, 4321e4327. Connors, N.C., Kofuji, P., 2006. Potassium channel Kir4.1 macromolecular complex in retinal glial cells. Glia 53, 124e131. Dalloz, C., Sarig, R., Fort, P., Yaffe, D., Bordais, A., Pannicke, T., Grosche, J., Mornet, D., Reichenbach, A., Sahel, J., Nudel, U., Rendon, A., 2003. Targeted inactivation of dystrophin gene product Dp71: phenotypic impact in mouse retina. Hum. Mol. Genet. 12, 1543e1554. Faude, F., Francke, M., Makarov, F., Schuck, J., Ga¨rtner, U., Reichelt, W., Wiedemann, P., Wolburg, H., Reichenbach, A., 2001. Experimental retinal detachment causes widespread and multilayered degeneration in rabbit retina. J. Neurocytol 30, 379e390. Fisher, S.K., Lewis, G.P., 2003. Mu¨ller cell and neuronal remodeling in retinal detachment and reattachment and their potential consequences for visual recovery: a review and reconsideration of recent data. Vision Res. 43, 887e897. Francke, M., Pannicke, T., Biedermann, B., Faude, F., Wiedemann, P., Reichenbach, A., Reichelt, W., 1997. Loss of inwardly rectifying potassium currents by human retinal glial cells in diseases of the eye. Glia 20, 210e218. Francke, M., Faude, F., Pannicke, T., Bringmann, A., Eckstein, P., Reichelt, W., Wiedemann, P., Reichenbach, A., 2001. Electrophysiology of rabbit Muller (glial) cells in experimental retinal detachment and PVR. Invest. Ophthalmol. Vis. Sci. 42, 1072e1079. Francke, M., Weick, M., Pannicke, T., Uckermann, O., Grosche, J., Goczalik, I., Milenkovic, I., Uhlmann, S., Faude, F., Wiedemann, P., Reichenbach, A., Bringmann, A., 2002. Up-regulation of extracellular ATP-induced Mu¨ller cell responses in a dispase model of proliferative vitreoretinopathy. Invest. Ophthalmol. Vis. Sci. 43, 870e881. Francke, M., Faude, F., Pannicke, T., Uckermann, O., Weick, M., Wolburg, H., Wiedemann, P., Reichenbach, A., Uhlmann, S., Bringmann, A., 2005. Glial cell-mediated spread of retinal degeneration during detachment: a hypothesis based upon studies in rabbits. Vision Res. 45, 2256e2267. Iandiev, I., Biedermann, B., Bringmann, A., Reichel, M.B., Reichenbach, A., Pannicke, T., 2006. Atypical gliosis in Mu¨ller cells of the slowly degenerating rds mutant mouse retina. Exp. Eye Res. 82, 449e457. Ishii, M., Horio, Y., Tada, Y., Hibino, H., Inanobe, A., Ito, M., Yamada, M., Gotow, T., Uchiyama, Y., Kurachi, Y., 1997. Expression and clustered distribution of an inwardly rectifying potassium channel, KAB-2/Kir4.1, on mammalian retinal Muller cell membrane: their regulation by insulin and laminin signals. J. Neurosci 17, 7725e7735. Ishii, M., Fujita, A., Iwai, K., Kusaka, S., Higashi, K., Inanobe, A., Hibino, H., Kurachi, Y., 2003. Differential expression and distribution of Kir5.1 and Kir4.1 inwardly rectifying Kþ channels in retina. Am. J. Physiol 285, C260eC267. Kofuji, P., Ceelen, P., Zahs, K.R., Surbeck, L.W., Lester, H.A., Newman, E.A., 2000. Genetic inactivation of an inwardly rectifying potassium channel (Kir4.1 subunit) in mice: phenotypic impact in retina. J. Neurosci 20, 5733e5740. Kofuji, P., Biedermann, B., Siddharthan, V., Raap, M., Iandiev, I., Milenkovic, I., Thomzig, A., Veh, R.W., Bringmann, A., Reichenbach, A., 2002. Kir potassium channel subunit expression in retinal glial cells: implications for spatial potassium buffering. Glia 39, 292e303. Kubo, Y., Baldwin, T.J., Jan, Y.N., Jan, L.Y., 1993. Primary structure and functional expression of a mouse inward rectifier potassium channel. Nature 362, 127e133. Marc, R.E., Murry, R.F., Fisher, S.K., Linberg, K.A., Lewis, G.P., 1998. Amino acid signatures in the detached cat retina. Invest. Ophthalmol. Vis. Sci. 39, 1694e1702.

313

Nagelhus, E.A., Horio, Y., Inanobe, A., Fujita, A., Haug, F.M., Nielsen, S., Kurachi, Y., Ottersen, O.P., 1999. Immunogold evidence suggests that coupling of Kþ siphoning and water transport in rat retinal Muller cells is mediated by a coenrichment of Kir4.1 and AQP4 in specific membrane domains. Glia 26, 47e54. Newman, E.A., 1987. Distribution of potassium conductance in mammalian Mu¨ller (glial) cells: a comparative study. J. Neurosci 7, 2423e2432. Newman, E.A., 1993. Inward-rectifying potassium channels in retinal glial (Mu¨ller) cells. J. Neurosci 13, 3333e3345. Newman, E.A., Reichenbach, A., 1996. The Mu¨ller cell: a functional element of the retina. Trends Neurosci 19, 307e317. Noel, G., Belda, M., Guadagno, E., Micoud, J., Klocker, N., Moukhles, H., 2005. Dystroglycan and Kir4.1 coclustering in retinal Mu¨ller glia is regulated by laminin-1 and requires the PDZ-ligand domain of Kir4.1. J. Neurochem 94, 691e702. Nork, T.M., Wallow, I.H.L., Sramek, S.J., Anderson, G., 1987. Mu¨ller’s cell involvement in proliferative diabetic retinopathy. Arch. Ophthalmol 105, 1424e1429. Osborne, N.N., Casson, R.J., Wood, J.P., Chidlow, G., Graham, M., Melena, J., 2004. Retinal ischemia: mechanisms of damage and potential therapeutic strategies. Prog. Retin. Eye Res. 23, 91e147. Pannicke, T., Iandiev, I., Uckermann, O., Biedermann, B., Kutzera, F., Wiedemann, P., Wolburg, H., Reichenbach, A., Bringmann, A., 2004. A potassium channel-linked mechanism of glial cell swelling in the postischemic retina. Mol. Cell. Neurosci 26, 493e502. Pannicke, T., Uckermann, O., Iandiev, I., Wiedemann, P., Reichenbach, A., Bringmann, A., 2005. Ocular inflammation alters swelling and membrane characteristics of rat Mu¨ller glial cells. J. Neuroimmunol 161, 145e154. Pannicke, T., Iandiev, I., Wurm, A., Uckermann, O., vom Hagen, F., Reichenbach, A., Wiedemann, P., Hammes, H.-P., Bringmann, A., 2006. Diabetes alters osmotic swelling characteristics and membrane conductance of glial cells in rat retina. Diabetes 55, 633e639. Pfaffl, M., 2001. A new mathematical model for relative quantification in realtime RTePCR. Nucleic Acids Res. 29, 2002e2007. Puwarawuttipanit, W., Bragg, A.D., Frydenlund, D.S., Mylonakou, M.N., Nagelhus, E.A., Peters, M.F., Kotchabhakdi, N., Adams, M.E., Froehner, S.C., Haug, F.M., Ottersen, O.P., Amiry-Moghaddam, M., 2006. Differential effect of alpha-syntrophin knockout on aquaporin-4 and Kir4.1 expression in retinal macroglial cells in mice. Neuroscience 137, 165e175. Raap, M., Biedermann, B., Braun, P., Milenkovic, I., Skatchkov, S.N., Bringmann, A., Reichenbach, A., 2002. Diversity of Kir channel subunit mRNA expressed by retinal glial cells of the guinea-pig. Neuroreport 13, 1037e1040. Rasmussen, R., 2001. Quantification on the LightCycler. In: Meuer, S., Wittwer, C., Nakagawara, K. (Eds.), Rapid Cycle Real-time PCR, Methods and Applications. Springer, Heidelberg, pp. 21e34. Reichenbach, A., Eberhardt, W., 1986. Intracellular recordings from isolated rabbit retinal Mu¨ller (glial) cells. Pflu¨gers Arch. 407, 348e353. Schnitzer, J., 1985. Distribution and immunoreactivity of glia in the retina of the rabbit. J. Comp. Neurol 240, 128e142. Takumi, T., Ishii, T., Horio, Y., Morishige, K., Takahashi, N., Yamada, M., Yamashita, T., Kiyami, H., Sohmiya, K., Nakanishi, S., Kurachi, Y., 1995. A novel ATP-dependent inward rectifier potassium channel expressed predominantly in glial cells. J. Biol. Chem. 270, 16339e16346. Yang, D., Pan, A., Swaminathan, A., Kumar, G., Hughes, B.A., 2003. Expression and localization of the inwardly rectifying potassium channel Kir7.1 in native bovine retinal pigment epithelium. Invest. Ophthalmol. Vis. Sci. 44, 3178e3185.