Retinal Colocalization and In Vitro Interaction of the Glutamate Receptor EAAT3 and the Serum- and Glucocorticoid-Inducible Kinase SGK1

Abstract

purpose. The serum- and glucocorticoid-inducible kinase SGK1 regulates several epithelial channels and transporters, the related protein kinase B (PKB) regulates glucose transport. SGK1 is expressed in the brain and could thus regulate glial and/or neuronal transport processes. The present study explores whether SGK1 is expressed in the retina and whether it regulates EAAT3, a Na⁺-coupled glutamate transporter. EAAT3 is expressed in retinal ganglion cells and accomplishes the clearance of glutamate from synaptic clefts.

methods. Immunohistochemistry was performed to test for retinal SGK1 expression. For functional analysis, cRNA encoding EAAT3 was injected into Xenopus oocytes with or without additional injection of wild-type SGK1, constitutively active ^S422DSGK1, inactive ^K127NSGK1, and/or constitutively active ^T308D,S473DPKB. Glutamate induced current (I_GLU) was taken as a measure for transport.

results. SGK1 is indeed expressed in several retinal cells including retinal ganglion cells where it is colocalized with EAAT3. In EAAT3-expressing Xenopus oocytes, glutamate-induced current was stimulated by coexpression of wild-type SGK1, constitutively active ^S422DSGK1, and constitutively active ^T308D,S473DPKB, but not by inactive ^K127NSGK1.

conclusions. SGK1 and EAAT3 are coexpressed in retinal neurons, and SGK1 serves to stimulate EAAT3. This function is shared by protein kinase B (PKB). The experiments reveal a novel mechanism regulating EAAT3, which may be essential for the function of the retinal ganglion cells.

The serum and glucocorticoid-inducible kinase SGK1 was originally cloned as a glucocorticoid-sensitive gene. ^{1 2 3} The human orthologue was cloned as a cell-volume–sensitive gene upregulated by osmotic and isotonic cell shrinkage. ⁴ More recent experiments disclosed the involvement of SGK1 in the regulation of a variety of channels and transporters, ⁵ such as the renal epithelial Na⁺ channel ENaC ^{6 7 8 9 10 11} ; the voltage-gated Na⁺ channel SCN5A ⁶ ; the K⁺ channels ROMK1, ¹² KCNE1/KCNQ1, ¹³ and Kv1.3 ^{14 15 16} ; the Na⁺/H⁺ exchanger NHE3 ¹⁷ ; and Na⁺/K⁺-ATPase. ^{18 19} As shown for the regulation of ENaC, SGK1 is at least partially effective through inhibition of neuronally expressed developmentally downregulated gene Nedd4-2, ^{20 21 22} a ubiquitin ligase expressed in a wide variety of tissues including the brain. ²³ SGK1 is related to PKB, which has been shown to be expressed in retinal tissue ²⁴ and is known to regulate the glucose transporter GLUT4. ^{25 26 27 28 29 30} Activation of SGK1 requires phosphorylation at Ser422, activation of PKB requires phosphorylation at Thr308 and Ser473. Replacement of those amino acids by aspartate leads to the respective constitutively active kinases ^S422DSGK1 ³¹ and ^T308D,S473DPKB. ³² Destruction of the catalytic subunit by replacement of the lysine at position 127 with asparagine leads to the inactive mutant ^K127NSGK1. ³¹ In vivo, SGK1 and PKB are activated through a signaling cascade involving phosphatidyl-inositol 3′ kinase (PI3-K) and phosphoinositide-dependent kinase PDK1. ³³ SGK1 ^{4 16} and PKB ^{32 34 35 36 37} are expressed in all human tissues studied thus far, including the brain. Among the stimulators of those kinases is presumably brain-derived neurotrophic factor (BDNF) which has been shown to activate PI3 kinase and PKB in retinal ganglion cells. ³⁸ The present study was performed to explore whether SGK1 is expressed in the retina. We showed strong expression of SGK1 in several neuronal cells, including the retinal ganglion cells. 

Additional studies were performed to gain some insight into the functional significance of SGK1. To this end, experiments were performed elucidating the influence of SGK1 on the glutamate transporter EAAT3 (EAAC1), a member of the Na⁺-coupled glutamate transporter family. ^{39 40 41 42 43 44 45 46 47} The carrier participates in the clearance of glutamate from the synaptic cleft. ^{48 49 50 51 52} It is expressed in bipolar, amacrine, and retinal ganglion cells. ⁵³ To test for a role of SGK1 and/or PKB in the regulation of EAAT3, cRNA encoding EAAT3 was injected, with or without cRNA encoding Nedd4-2 and/or constitutively active ^S422DSGK1, inactive ^K127NSGK1, or constitutively active ^T308D,S473DPKB, into Xenopus oocytes. The glutamate-induced currents were taken as a measure of glutamate transport. 

To explore in vivo regulation of glutamate transport in retinal ganglion cells, glutamate induced currents in the presence of glutamate receptor blockers were determined by patch clamp in retinal ganglion cells from BDNF-deficient (BDNF ^−/−) and in wild-type (BDNF ^+/+) mice. 

Materials and Methods

Animal Experimentation

All experiments performed in this study were in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and with the guidelines of the ethics committees of the University of Tübingen. 

For immunohistochemistry, retinas from adult (postnatal day [P]92) Brown Norway (BN) rats reared in our animal facilities were used in the study. The animals were kept under a 12-hour light-dark cycle, with an illumination of 10 lux. Tissue samples were obtained between 9 AM and 5 PM, during the light phase. 

Glutamate transporter currents in retinal ganglion cells were studied in homozygous BDNF-deficient (BDNF ^−/−) and wild-type (BDNF ^+/+) mice ⁵⁴ at P3 and P4. The exact genotype of each tested mouse was determined by Southern blot analysis. 

Immunohistochemistry

The animals were killed with CO₂ and decapitated. Their eyes were enucleated and the anterior poles removed. The remaining posterior eyecups were fixed in 4% paraformaldehyde (PFA) in phosphate buffer (PB; 0.2 M NaH₂PO₄/Na₂HPO₄; pH 7.4) for 30 minutes at 4°C. After being washed in PB, the eyecups were cryoprotected by immersion in 30% sucrose in PB overnight at 4°C and embedded (Tissue Tek; Leica, Nussloch, Germany). Vertical cryosections of 10 μm thickness were collected on gelatin-coated glass slides, dried for 3 hours and stored at −80°C until use. 

To reduce background staining, sections were incubated for 30 minutes with 20% normal goat serum (NGS; Sigma-Aldrich, Taufkirchen, Germany) in phosphate-buffered saline (PBS) containing 0.1% Triton X-100 (PBST). For immunohistochemical detection of SGK1, a specific polyclonal antiserum raised in rabbit was used. ¹⁶ A monoclonal mouse anti-EAAT3 antibody was purchased from Chemicon (Temecula, CA). The antibodies were diluted in PBST (1:250 and 1:100, respectively) and used to incubate sections overnight at 4°C. After they were washed with PBS, the sections were incubated for 1 hour with the appropriate secondary antibodies (goat anti-rabbit or goat anti-mouse; dilution 1:200; Sigma-Aldrich) conjugated to Alexa Fluor 488 or conjugated to Cy3 (Molecular Probes, Eugene, OR). Double-labeling studies were performed by simultaneously applying the primary antibodies followed by a mixture of secondary antibodies. Labeled sections were washed in PBS and coverslipped with glycerol. For determination of the specificity of the antibody-reaction, negative control were performed by omitting the primary antibody. 

All histologic examinations and photodocumentation were performed by microscope (model AX70; Olympus, Tokyo, Japan). Photographs were digitized and adjusted for brightness and contrast (Photoshop; Mountain View, CA). 

Construction of Xenopus laevis Expression Vectors

cRNA encoding human EAAT3, ^{39 40} human Nedd4-2, ²⁰ wild-type SGK1, ⁴ constitutively active human ^S422DSGK1, ³¹ inactive ^K127NSGK1, ³¹ and constitutively active human ^T308D,S473DPKB ³² were synthesized as described. ⁵⁵  

Two-Electrode Voltage Clamp

Dissection of Xenopus laevis ovaries and collection and handling of the oocytes have been described in detail elsewhere. ⁵⁵ Oocytes were injected with 7.5 ng ^S422DSGK1, ^K127NSGK1, ^T308D,S473DPKB cRNA, and/or 5 ng Nedd4-2 cRNA on the first day after preparation of the oocytes and with 10 ng EAAT3 cRNA on the second day after preparation. Inactive ^K127NSGK1 served as the control for the active kinases. When no cRNA was injected, water was injected as a control to avoid any bias by the injection itself. All experiments were performed at room temperature 4 days after the second injection. Two-electrode voltage-clamp recordings were performed at a holding potential of −60 mV. The data were filtered at 10 Hz, and recorded with an A/D-D/A converter (MacLab; ADInstruments, Castle Hill, Australia) and software for data acquisition and analysis (ADInstruments). The control solution (superfusate/ND96) contained (in mM) 96 NaCl, 2 KCl, 1 CaCl₂, 1 MgCl₂, and 5 HEPES [pH 7.4]. The final solutions were titrated to the pH indicated using HCl or NaOH. The flow rate of the superfusion was 20 mL/min, and a complete exchange of the bath solution was reached within approximately 10 seconds. 

Patch Clamp

To determine glutamate transporter currents in retinal ganglion cells from homozygous BDNF-deficient (BDNF ^−/−) and wild-type mice (BDNF ^+/+) at P3 and P4, postnatal ^C57BL6BDNF mice pups were killed by decapitation. After removal of the eyeballs, retinas were dissected, cut into strips, and kept at room temperature in continuously bubbled (95% O₂, 5% CO₂) bicarbonate-buffered ACSF, containing (in mM) 125 NaCl, 3.5 KCl, 1 MgCl₂, 25 NaHCO₃, 1.25 NaH₂PO₄, 2 CaCl₂, and 25 glucose, at pH 7.4, adjusted with 1 M NaOH. Recordings were made between 2 and 5 hours after retinal dissection. All measurements were performed at a temperature of 23°C to 25°C. Cells were chosen for recording according to position in the ganglion cell layer and soma size. Only large-diameter ganglion cells were selected, because type I retinal ganglion cells have been shown to have much larger cell somata than displaced amacrine cells and other types of retinal ganglion cells. Therefore, we assume that we mainly recorded from this ganglion cell type. 

For the isolation of glutamate transporter currents, ACSF was replaced with a solution with lowered calcium, containing (in mM): 125 NaCl, 3.5 KCl, 25 NaHCO₃, 1.25 NaH₂PO₄, 0.2 CaCl₂, 1 MgCl₂, and 25 glucose [pH 7.4, adjusted with 1 M NaOH]. Both solutions were held at room temperature and continuously bubbled with carbogen during the whole experiment. The pipette solution (ic) contained (in mM): 140 KCl₂, 2 MgCl₂, 10 EGTA, 4 Na₂-ATP, 0.44 Na-GTP, and 10 HEPES [pH 7.2, adjusted with 1 M KOH]. The inhibition of different glutamate current types in rat retinal ganglion cells was based on pharmacological properties. Drugs were applied to the bath in the following concentrations CNQX (25 μM), DAP-5 (10 μM), MCPG (500 μM). Remaining spontaneous activity was blocked by the addition of strychnine (200 μM) and picrotoxin (200 μM). Glutamate (1 mM) was applied by a six-barrel superfusion pipette. All drugs were purchased from Sigma-Aldrich (Deisenhofen, Germany) and dissolved in bath solution just before use. Drug solutions were bubbled with 95% O₂-5% CO₂ at room temperature immediately before application to retinal slices. 

For electrophysiological recordings, wholemount strips were transferred into the recording chamber on a microscope stage (Axioskop; Carl Zeiss Meditec, Oberkochen, Germany) and fixed with a small grid made of fine nylon strings tightened between a U-shaped platinum wire. The recording chamber was superfused with 1.5- to 2-mL/min ACSF. Patch pipettes were pulled out of borosilicate capillaries. Pipettes were filled with intracellular solution (ic). When filled, they displayed resistances of 2.0 to 3.5 MΩ. Recordings were performed with an EPC-7 amplifier (Heka Elektronik, Lambrecht, Germany), using optimal series resistance compensation. Series resistance compensation was usually 60% to 80%. Cells were selected for study if they had series resistances less than 15 MΩ. Whole-cell currents were recorded with sampling frequencies of 100 and 5 kHz. Data were not corrected for the liquid junction. Cells were normally kept at a holding potential of −60 mV. 

A commercial software program (Clampex, ver. 8.1.0; Axon Instruments, Union City, CA) was used for data acquisition and analysis, and another (Sigmaplot; SPSS Inc., Chicago, IL) was used for curve fitting and plotting. Time constants of decay (τ) were determined by fitting the single exponential function A _n · exp^{(tK)/ n} + C to the data, where A is the amplitude relative to offset evaluated at the start of the fit region (n), C is the steady state asymptote, and (tK) is the time, set to zero at the beginning (K) of the fit region. Fitting procedures were performed on computer (Clampfit; Axon Instruments) and IGOR (Wavemetrics, Lake Oswego, OR) software routines. The amplitude was determined by measuring the time from the beginning of the current until the peak current was reached. 

Statistical Analysis

Data are provided as arithmetic means ± SEM, and n represents the number of oocytes investigated. All experiments were repeated in at least three batches of oocytes. In all repetitions, qualitatively similar data were obtained. All data were tested for significance by ANOVA, and only results with P < 0.05 were considered statistically significant. 

Results

SGK1 Expression and Colocalization with EAAT3 in Rat Retina

To examine cell-specific protein localization, we performed immunohistochemical stainings in adult rat retina sections using our polyclonal anti-SGK1 antiserum. ¹⁶ By using this antiserum, we observed a strong staining of most neurons in the ganglion cell layer (GCL; 1 , arrows). Individual cell bodies of different sizes were labeled with a particular strong immunoreaction in cells with a diameter larger than 12 μm. The fiber layer built by the ganglion cell axons did not show any SGK1 immunoreactivity. As approximately half of the cells in the GCL of the rodent retina are displaced amacrine cells, both ganglion cells and displaced amacrine cells showed a positive SGK1 immunoreactivity. 

In addition to the cells in the ganglion cell layer, staining was found in the cell bodies in the inner nuclear layer (INL). A first subset of cells in the innermost row of the INL along the border to the inner plexiform layer (IPL) was immunoreactive (1 , triangle and double arrows). Because of their position, the cells can be addressed as amacrine cells. An immunoreaction was also visible in three narrow, stratified bands in the IPL 1 . Band one was running in the outermost IPL along the border to the INL and band two at 25% and band three at 50% depth between the INL and GCL. However, we do not yet know which of the cell types from the GCL and the INL are connected to which of the three different sublayers in the IPL. No staining was present in the outer plexiform layer (OPL). No staining was visible when the SGK1 antibody was omitted (data not shown). 

Staining with an EAAT3-specific antibody revealed expression of the transporter in apparently all cells of the GCL 1 . Arrows denote prominent immunoreactivity of the GCL composed of ganglion and displaced amacrine cells. EAAT3 expression was also found in the cell bodies in the inner nuclear layer. The INL shows partial immunoreactivity with most intensive staining of a subset of cells in close proximity to the inner plexiform layer (1 , triangle and double arrows). These cells can be assigned as amacrine cells. Antibody staining occurred also in the OPL 1 and in the IPL. EAAT3 immunoreactivity was more uniformly distributed within the IPL, with an alleviated reaction in the inner part of the IPL toward the GCL 1 . Control experiments omitting the first antibody resulted in a complete absence of immunoreactivity over the entire retinal tissue (data not shown). 

Double labeling with both the SGK1 and the EAAT3 antibody 1 , demonstrated an overlap of SGK1 and EAAT3 expression in amacrine cells in the inner row of the INL (1 , triangle) and in cells in the GCL (1 , arrows). Besides colocalization, cells without a clear sign of an overlap were present in the INL and in the GCL (1 , double arrows). Whereas the OPL did not show any coexpression, a colocalization within the INL is very likely, even though a distinct layering for the EAAT3 antibody corresponding to that of SGK1 was missing. However, both antibodies did not stain (or had at least a markedly reduced immunoreactivity in) the inner part of the IPL. Moreover, the border of the internal IPL staining for EAAT3 corresponded with the one of the innermost SGK1 band 1 . 

Regulation of EAAT3 by SGK1

In Xenopus oocytes expressing the glutamate transporter EAAT3, but not in water injected oocytes, the addition of 1 mM glutamate to the bath induced an inward current (I_GLU) approaching −21.6 ± 1.4 nA (n = 30) at a holding potential of −60 mV. As shown in 2 , coexpression of wild-type SGK1 significantly increased the EAAT3 mediated current by 55.1% ± 6.2% (n = 26). 

The effect of wild-type SGK1 was mimicked by the constitutively active ^S422DSGK1 3 , which increased the EAAT3 mediated current by 55.0% ± 8.2% (n = 13). In contrast, the inactive ^K127NSGK1 did not significantly modify I_GLU (+4.8% ± 7.9%, n = 13). Similar to SGK1 and ^S422DSGK1, the constitutively active ^T308D,S473DPKB stimulated I_GLU in EAAT3-expressing oocytes 3 . Coexpression of ^T308D,S473DPKB increased I_GLU by 75.83% ± 5.8% (n = 26). 

I_GLU was not significantly altered (−5.2% ± 3.1%, n = 11) after coexpression of Nedd4-2 3 . Moreover, the effects of SGK1, ^S422DSGK1, and ^T308D,S473DPKB were not significantly modified by coexpression of Nedd4-2 3 . In Xenopus oocytes coexpressing EAAT3 together with SGK1, ^S422DSGK1, or ^T308D,S473DPKB and Nedd4-2, I_GLU was not significantly different from I_GLU in oocytes expressing EAAT3 with the respective kinases alone but significantly larger than I_GLU in Xenopus oocytes expressing EAAT3 alone. 

As illustrated in 4 addition of glutamate to mouse retinal ganglion cells in the presence of the glutamate receptor blockers CNQX (25 μM), DAP-5 (10 μM), and MCPG (500 μM), as well as strychnine (200 μM) and picrotoxin (200 μM), induced a current that was most likely generated by electrogenic glutamate uptake. The current was significantly smaller in retinal ganglion cells from BDNF-deficient mice (BDNF ^−/−) compared with the current in retinal ganglion cells from wild-type mice (BDNF ^+/+). 

Discussion

EAAT3 encodes for a carrier protein mediating the electrogenic, Na⁺-coupled transport of glutamate. ^{39 40} As a transport of negatively charged glutamate is coupled to the translocation of 3 Na⁺ ions, transport of glutamate through EAAT3 is coupled to the translocation of charge. ^{39 40 56 57} The current generated by addition of substrate is a measure of glutamate transport. Xenopus oocytes do not express endogenous electrogenic glutamate transport—that is, the current is solely generated by the heterologously expressed transport protein. 

The present observations reveal a novel mechanism for regulation of EAAT3—that is, the stimulation by SGK1 and the PKB. The kinases stimulate EAAT3 in the absence of heterologously expressed Nedd4-2, and their effect is not modified by the additional expression of Nedd4-2. Moreover, Nedd4-2 does not significantly modify EAAT3 activity. Thus, in contrast to what has previously been shown for the regulation by SGK1 of the voltage gated cardiac Na⁺ channel ⁵⁸ and the epithelial Na⁺ channel ENaC, ^{20 21} the kinases do not regulate EAAT3 through inhibition of Nedd4-2 in this system. Other signaling molecules involved in the regulation of ENaC include the small G protein K-Ras 2A and the channel activating protein (CAP)-1. ^{59 60} The regulation of renal outer medullary K⁺ channel ROMK1 ¹² and of the Na⁺/H⁺ exchanger NHE3 by SGK1 involves the Na⁺/H⁺ exchanger regulating factor NHERF2. Moreover, regulation of ROMK1 involves direct phosphorylation of the channel protein. ⁶¹ Thus, several candidate mechanisms may mediate the effect of SGK1 on EAAT3. 

The effects of SGK1 and PKB on the glutamate transporter EAAT3 presumably participate in the clearance of glutamate from the extracellular space and thus in the termination of the excitatory signal. As shown in the present study, SGK1 is expressed in several retinal cells, including retinal ganglion cells, and PKB has been shown to be expressed in retinal tissue. ²⁴ The decreased glutamate induced currents in retinal ganglion cells from BDNF-deficient mice points to a role of BDNF in the regulation of glutamate transport. Because BDNF signals through PI3 kinase and PKB in retinal ganglion cells, ³⁸ it is expected similarly to activate SGK1. The two kinases could thus contribute to the effects of BDNF in those cells. 

In theory, stimulation of glutamate transport could provide some protection against excessive activation by glutamate. Experimental evidence points to a role for glutamate in apoptosis after nerve crush ⁶² and in mutant quails, an animal model for glaucoma, glutamate concentrations increase immediately before cell death. ³⁸ External application of glutamate is known to trigger retinal ganglion cell death, ^{63 64 65 66 67} at least partially, by increase of cytosolic Ca²⁺ activity, ⁶⁸ activation of caspases, ^{69 70} and subsequent apoptosis. ⁷¹ Conversely, PKB activity has been shown to correlate with retinal cell survival. ^{24 62} Accordingly, dephosphorylation and thus deactivation of PKB by inhibition of PI3 kinase triggers apoptosis of retinal cells. ²⁴ Notably, this maneuver not only disrupts PKB activity but similarly downregulates SGK1 activity. The effect is not reversed by activation of PKA and subsequent phosphorylation of Bad. Thus, the authors concluded that the antiapoptotic effect of PKB was not due to phosphorylation of Bad. ²⁴ In theory, the mechanism could involve regulation of glutamate uptake by PKB and SGK1. Nothing is known, however, about the specific role of EAAT3 in the survival of retinal ganglion cells. Antisense knockdown of EAAT3 in mice did not lead to severe neurodegeneration but, besides dicarboxylic amino aciduria, caused some subtle behavioral abnormalities and seizures. ⁷² Dicarboxylic amino aciduria in humans may be similarly paralleled by neurologic abnormalities. ⁷³ The lack of severe neurodegeneration could be due to glutamate uptake through other glutamate transporters which may be similarly sensitive to SGK1 and PKB. ⁵⁸ Taken together, presently available evidence does not allow the speculation that SGK1- or PKB-dependent regulation of EAAT3 participates in the regulation of retinal cell survival. 

In conclusion, the present observations demonstrated that SGK1 is expressed in retinal neurons and disclosed a novel function of this kinase—that is, the regulation of the glutamate transporter EAAT3. The effect of SGK1 is shared by PKB. The mechanism is likely to participate in the regulation of glutamate clearance from the extracellular space. 

 Supported by the Deutsche Forschungsgemeinschaft Grants La 315/4-4, La 315/5-1, and Br 1318/2-4; and the Bundesministerium für Bildung, Wissenschaft, Forschung, und Technologie (Center for Interdisciplinary Clinical Research) Grant 01 KS 9602.

 Submitted for publication January 22, 2003; revised June 30 and December 16, 2003, and January 15, 2004; accepted January 15, 2004.

 Disclosure: R. Schniepp, None; K. Kohler, None; T. Ladewig, None; E. Guenther, None; G. Henke, None; M. Palmada, None; C. Boehmer, None; J.D Rothstein, None; S. Bröer, None; F. Lang, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Figure 1.

 

SGK1 and EAAT3 protein expression in neurons of the rat retina: radial sections of adult rat retina simultaneously labeled with SGK1 (A) and EAAT3 (B) antibodies and their colocalization (C). (A) Cellular localization of SGK1 protein: cell bodies are marked in the GCL (arrows) and in the inner row of the INL (triangle; small double arrows). A positive immunoreaction is present in three distinct bands in the IPL. (B) Cellular localization of EAAT3 protein. Similar to SGK1 immunostaining, cell bodies in the GCL (arrows) and in the inner row of the INL are marked (triangle, small double arrows in the INL). Both plexiform layers, OPL and IPL, showed an EAAT3-positive reaction. In contrast to the ribbon-like SGK1 distribution (A) EAAT3 was more uniformly distributed over the IPL. (C) Merged images of (A) and (B). Colocalization of SGK1 and EAAT3 is indicated by a yellow stain and was found in amacrine cells (triangle) and in most of the cells in the GCL (arrows). Cells without a clear sign for a colocalization of both proteins were present in the INL and GCL (A, B, C: small double arrows). Whereas the OPL did not show any coexpression, colocalization in the IPL is very likely, even though a distinct layering for the EAAT3 antibody corresponding to that of SGK1 is missing. Scale bar, 25 μm.

Figure 2.

 

EAAT3 transport activity was upregulated by SGK1. (A) Original tracings. Solid rectangles: administration of 1 mM glutamate. (B) Arithmetic means ± SEM. *Statistically significant compared with currents in Xenopus oocytes expressing EAAT3+H₂O.

Figure 3.

 

(A) EAAT3 transport activity was upregulated by the constitutively active ^S422DSGK but not by the inactive mutant ^K127NSGK1. Similar to SGK1, the constitutively active ^T308D,S473DPKB increased EAAT3 currents. (B) The ubiquitin ligase Nedd4-2 did not modify EAAT3 activity. On coexpression of Nedd4-2, glutamate-induced currents remained unchanged compared with the expression of EAAT3 alone. Moreover, the stimulating effect of the protein kinases SGK1, ^S422DSGK1, and ^T308D,S473DPKB on EAAT3 were not significantly different in the presence of Nedd4-2 than with expression of the protein kinases alone with EAAT3. Arithmetic means ± SEM. *Statistically significant compared with currents in Xenopus oocytes expressing EAAT3+H₂O.

Figure 4.

 

Blunted glutamate-induced current in retinal ganglion cells from BDNF-deficient mice (BDNF^−/−) and wild-type (BDNF^+/+) mice. The glutamate (1 mM)-induced currents were determined in retinal ganglion cells in slices of retinal tissue under blockage of glutamate receptors. (A) Original tracings; (B) arithmetic means ± SEM (n = 4) of the time constants (top) and maximum currents (bottom). *Statistically significant value in (BDNF^−/−) compared with wild-type (BDNF^+/+) mice.

 

The authors thank Sir Philip Cohen for providing the constructs of active ^S422DSGK1, inactive ^K127NSGK1, and constitutively active ^T308D,S473DPKB, and Birgitta Noll, Sylvia Bolz, and Gudrun Härer for expert technical assistance.