Glutamate Induces Rapid Upregulation of Astrocyte Glutamate Transport and Cell-Surface Expression of GLAST

MATERIALS AND METHODS

Immunopure immobilized monomeric avidin and sulfo-NHS-biotin were purchased from Pierce (Rockford, IL). Genistein, CF-109203X,trans-(±)-1-amino-1,3-cyclopentanedicarboxylic acid (t-ACPD),l(+)-2-amino-4-phosphonobutyric acid (l-AP4), (S)-4-carboxyphenylglycine (4-CPG), and 1-(5-isoquinolinesulfonyl)-2-methylpiperazine (H7) were purchased from RBI (Natick, MA); 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX),l-trans-pyrrolidine-2,4-dicarboxylate (t-PDC), and α-methyl-4-carboxyphenylglycine (MCPG) were purchased from Tocris (Ballwin, MO); and 1,2-bis(2-aminophenoxy)ethane-N,N,N,N-tetra-acetic acid acetoxymethyl ester (BAPTA AM) was purchased from Molecular Probes (Eugene, OR). All other reagents were obtained from Sigma (St. Louis, MO) except where noted.

Cell culture preparation. Primary murine cortical astrocyte cultures were prepared as described previously (Swanson and Seid, 1998). Tissue harvest was performed in accordance with National Institutes of Health guidelines in a manner that minimized animal suffering. In brief, cortices were harvested from 1-d-old mice (Simonsen, Gilroy, CA) under deep isofluorane anesthesia. The cortices were dissociated in papain/DNase, plated in 24-well tissue culture plates in Eagle's MEM containing 10% fetal bovine serum (FBS) (Hyclone, Ogden, UT), and 2 mm glutamine, and maintained at 37°C in a 5% CO₂ incubator. At confluence (day 12–15), the cultures were treated for 48 hr with 10 μmcytosine arabinoside to prevent proliferation of other cell types, and the medium was replaced with MEM containing 2 mm glutamine, 3% FBS, and 0.15 mm dibutyryl cAMP to induce differentiation (Sensenbrenner et al., 1980; Swanson et al., 1997). The astrocyte cultures formed a confluent layer of process-bearing, glial fibrillary acidic protein (GFAP)-positive cells. Each study was repeated on cells from at least two different batches of astrocyte cultures at 28–35 d in vitro.

Experimental procedures. Incubations and uptake assays were performed in room air, 37°C, using a balanced salt solution (BSS) containing (in mm): NaCl 135, KCl 3.1, CaCl₂ 1.2, MgSO₄ 1.2, KH₂PO₄ 0.5, PIPES 5, and glucose 2. pH was adjusted to 7.2 with NaOH. Osmolarity was measured with a vapor pressure osmometer (Wescor, Ogden, UT) and adjusted to 285–315 mOsm with water or NaCl if needed. Na⁺-free media was prepared by replacing NaCl with choline chloride and NaOH withN-methyl-d-glucamine. Ca²⁺-deficient medium was prepared by omitting CaCl₂. All reagents were diluted into BSS from iso-osmolar stocks prepared in BSS or Na⁺-free BSS.

Glutamate uptake was determined after preincubation with glutamate or other agents. Preincubations were initiated 6 min after washing the cultures into 300 μl of BSS. The preincubations were terminated by double washes in BSS. The cultures remained in BSS at 37°C until initiation of a glutamate uptake assay (the “wash period”). The preincubation period was 60 min, and the wash period was 6 min except where otherwise noted.

Glutamate uptake assays were initiated by adding 0.167 μCi/ml¹⁴C-glutamate (American Radiochemicals, St. Louis, MO) plus 100 μm unlabeled glutamate to the culture wells in a final volume of 300 μl. Uptake was terminated after 7 min by two ice-cold washes with 500 μl BSS followed by immediate lysis in ice-cold 0.1N NaOH/0.01% lauryl sulfate. Aliquots of lysates were taken for scintillation counting and protein determinations. In separate experiments, uptake was confirmed to be linear through 10 min incubation (data not shown).

In some studies, glutamate uptake was assessed by HPLC measurements of medium glutamate concentrations. Glutamate (100 μm) was added to the wells in a final volume of 300 μl, and the medium was harvested 7 min later. Glutamate concentrations were determined by reversed-phase HPLC of the orthophthaldialdehyde derivatives using a fluorescence detector. Peak areas were measured with Nelson Analytical software (Norwalk, CT) and calibrated against BSS samples with known glutamate concentrations.

Biotinylation. Biotinylation of cell surface proteins was performed as described by Qian et al. (1997) and Davis et al. (1998)with some modifications. After a 1 hr preincubation period and a 6 min BSS wash period, the cultures were rinsed twice with PBS/Ca-Mg containing (in mm): 138 NaCl, 2.7 KCl, 1.5 KH₂PO₄, 9.6 Na₂HPO₄, 1 MgCl2, and 0.1 CaCl₂, pH 7.3. Cultures were then incubated in sulfo-NHS-biotin solution (1 mg/ml in with PBS/Ca–Mg) for 20 min at 4°C. Biotinylation was terminated by washing twice in a quenching solution of PBS/Ca–Mg in which there was an equimolar substitution of 100 mm glycine for NaCl (to maintain 300 mOsm). This was followed by an additional 45 min incubation in the quenching solution at 4°C. Quenching solution was removed, and the cells were lysed with 100 μl/well of radioimmunoprecipitation assay (RIPA) buffer with protease inhibitors (100 mm Tris-HCl, pH 7.4, 150 mm NaCl, 1 mm EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mm iodoacetamide, 1 μg/ml leupeptin, 5 μg/ml aprotinin, and 250 μmphenylmethylsulfonyl fluoride) for 1 hr at 4°C with vigorous shaking. The lysates were centrifuged at 14,000 × g for 15 min at 4°C. Three hundred microliters of the supernatant were taken for Western analysis as the whole cell fraction. The rest of the supernatant (600 μl) was incubated with 300 μl avidin bead suspension for 1 hr at room temperature with gentle shaking. The avidin–lysate solution was then centrifuged for 15 min at 14,000 × g, and the supernatant was taken for Western analysis as the intracellular fraction. The pellet was washed four times with 1 ml RIPA buffer and resuspended in 300 μl Laemmli buffer (62.4 mm Tris-HCl, pH 7, 2% SDS, 20% glycerol, and 5% 2-mercaptethanol) for 1 hr with gentle shaking at room temperature. After centrifugation for 15 min at 14,000 × g, the supernatant was taken for Western analysis as the biotinylated (plasma membrane) fraction.

For Western analysis, the protein samples were loaded and run on a 10% SDS polyacrylamide gel. Samples were electrophoretically transferred onto polyvinylidene fluoride membranes (Boehringer Mannheim, Indianapolis, IN) and subsequently placed in blocking solution (5% skim milk/1% bovine serum albumin in 0.1 m phosphate buffer) for 1 hr at room temperature. The membranes were then transferred to a blocking solution containing polyclonal anti-actin antibody diluted 1:100 plus either polyclonal anti-GLAST antibody (0.2 ng/ml) or polyclonal anti-GLT-1 antibody (0.02 ng/ml) for 12 hr at 22°C. [The affinity-purified antibodies to GLAST and GLT-1 were kindly provided by Dr. Jeffrey Rothstein, Johns Hopkins University, and have been previously characterized (Rothstein et al., 1994; Swanson et al., 1997; Swanson and Seid, 1998)]. The membranes were washed three times in 0.1 m phosphate buffer containing 0.1% Tween 20 (PB-T) and placed in blocking solution containing a peroxidase-labeled anti-rabbit IgG antibody (Boehringer Mannheim), diluted 1:500, for 1 hr. After three additional washes in PB-T, the resulting peroxidase signal was detected using 3,3′-diaminobenzidene. The resulting bands were digitized, and densitometry was performed using the NIH Image program. The signal for each lane was calculated by summing the (area × (density-background)) measurements of the discrete monomer and multimer bands produced by GLAST and GLT-1.

Statistics. Statistical differences were determined by Student's t test for two-group comparisons, or by one-way ANOVA with the Tukey test for multiple comparisons among more than two groups, or by Dunnett's test when multiple groups were compared against a single control group.

RESULTS

DISCUSSION

The primary finding reported here is that glutamate can induce a rapid upregulation of astrocyte glutamate uptake. The studies further suggest that the glutamate-induced upregulation is signaled by increased activity of the GLAST (EAAT1) glutamate transporter subtype, and that the increase in uptake is mediated by increased cell-surface expression of GLAST. Additionally, a role for the actin cytoskeleton in this process is suggested by the finding that the increase in uptake activity is inhibited by cytochalasins.

Several factors support the conclusion that the glutamate-induced stimulation of glutamate uptake is triggered by increased activity of the glutamate transporters. The glutamate effect was mimicked byd-aspartate, which is also a transporter substrate, stimulation was blocked if the glutamate preincubation was performed in a Na⁺-free medium, and stimulation was not blocked by glutamate receptor antagonists that are not substrates for the transporters. In addition, maximal stimulation of uptake was observed only at glutamate concentrations that saturate glutamate transport, and these concentrations far exceed known receptor affinities for glutamate.

The membrane biotinylation studies provide evidence that the increase in uptake is mediated by transporter translocation to (or decreased removal from) the cell membrane. The observed increase in transportV_max with little or no change in glutamate K_m is consistent with this mechanism. The time course of the glutamate stimulation also supports a trafficking mechanism, because the 12–21 min interval between onset of glutamate preincubation and the resultant increase in uptake activity shown in Figure 1B is too fast to involve de novo synthesis of new glutamate transporters but is typical of the time interval required for changes attributable to membrane trafficking (Davis et al., 1998; Bernstein and Quick, 1999; Lissin et al., 1999).

GLAST and GLT-1 are the only glutamate transporter subtypes known to be expressed by mammalian forebrain astrocytes, both in culture andin situ (Rothstein et al., 1994; Swanson et al., 1997;Gegelashvili and Schousboe, 1998). In the present study, glutamate-induced stimulation of glutamate uptake is attributed to GLAST because the GLT-1 selective inhibitor dihydrokainate had negligible effect, and because GLAST but not GLT-1 exhibited increased cell surface expression. It is possible that different, as yet unrecognized transporters contribute to the stimulated uptake, but this seems unlikely because the magnitude of the increase is large (∼100%).

Glutamate has previously been shown to regulate glutamate transport in other ways. Expression of GLAST protein in cultured astrocytes is stimulated by prolonged exposure (7 d) to glutamate or kainate (Gegelashvili et al., 1996), and this increase is accompanied by an increase in transport V_max. Glutamate preincubation has also been reported to reduce extracellular glutamate concentrations in astrocyte cultures (Ye and Sontheimer, 1999). Interestingly, this effect displays a time course similar to that observed in the present studies, occurs in response to metabotropic glutamate receptor agonists as well as to glutamate itself, but is not blocked by metabotropic receptor antagonists. Because many of the metabotropic agonists can also serve as substrates for the transporters (Harris et al., 1987; Ye and Sontheimer, 1998), it is possible that this effect is also mediated by a substrate-stimulated increase in transporter activity.

Although the present studies suggest an upstream signaling mechanism by which increased extracellular glutamate concentrations may trigger the increase in glutamate uptake signal and a downstream effector mechanism by which the increase is accomplished, they do not establish an intervening signal transduction pathway. Previous studies have identified mechanisms controlling membrane trafficking of a predominately neuronal glutamate transporter, EAAC1 (EAAT3). In C6 glioma cells expressing EAAC1, PKC activation by phorbol ester leads to transporter movement from the intracellular compartment to the cell surface, whereas the PI3K inhibitor wortmannin decreased the membrane EAAC1 expression both in C6 glioma cells (Davis et al., 1998) and in neurons (Munir et al., 1998). Less is known about transporter trafficking in astrocytes, but it has been reported that membrane expression of GLAST, but not GLT-1, is inhibited by phorbol ester (Correale et al., 1998). In the present study neither the PKC inhibitor GF-109203X nor the PI3K inhibitor wortmannin affected glutamate-induced stimulation of uptake, and various other interventions also failed to affect stimulation of uptake (Tables 1, 2). However, translocation of membrane proteins does not necessarily require signal transduction pathways. G-protein-coupled receptors (Grady et al., 1997) and ionotropic glutamate receptors (Lissin et al., 1999) appear to involve the conformation-dependent interaction of the proteins with cytoplasmic regulatory proteins without intervening signal transduction, and this raises the possibility of a similar mechanism for glutamate transporters. Alternatively, the glutamate transporters themselves may serve as signal transducing units as proposed by Gegelashvili and Schousboe (1998), who noted that the third intracellular domains of GLAST, GLT-1, and EAAC1 each contain a motif similar to that of the IGF-II and α-adrenergic receptors that bind Gα-subunits of G-proteins.

Substrate-induced upregulation of transport activity mediated by membrane translocation of transporters, as described here, has been reported recently in other systems. The binding of substrates to the 5-HT transporter can rapidly increase 5-HT uptake by inhibiting PKC-dependent internalization of the 5-HT transporters (Qian et al., 1997; Ramamoorthy and Blakely, 1999). A substrate-induced rapid increase in GABA uptake was also observed in primary hippocampal cultures and in a cell line expressing the rat brain GABA transporter GAT1 (Bernstein and Quick, 1999), with the increase in GABA uptake correlating with increased membrane GAT1 expression. PKC is also involved in the regulation of GAT1 activity and trafficking (Quick et al., 1997), but in both the 5-HT and GABAergic systems the signals linking increased substrate exposure to PKC activity remain unknown.

The inhibitory effects of cytochalasin B and cytochalasin D on glutamate-stimulated glutamate uptake suggest that the actin cytoskeleton is important for either signal transduction or transporter translocation in this process. The cytochalasins impair actin polymerization and promote disruption of the actin cytoskeleton (MacLean-Fletcher and Pollard, 1980). Cytochalasin B has, in addition to effects on actin polymerization, a direct inhibitory effect on glucose transport (Rampal et al., 1980), but cytochalasin D does not have any known direct effect on glucose transporters or other transporter function.

The actin cytoskeleton has been linked to intracellular protein trafficking, transporter function, and signal transduction in several systems (Mills and Mandel 1994; Tsakiridis et al., 1994; Lamaze et al., 1997; Wang et al., 1998). Cytochalasin D inhibits insulin-stimulated glucose uptake and prevents insulin-induced trafficking of glucose transporters to the plasma membrane in L6 rat skeletal muscle cells (Tsakiridis et al., 1994) and 3T3-L1 adipocytes (Wang et al., 1998). In this system the actin filaments are thought to facilitate the insulin-induced relocation of PI3K to the intracellular glucose transporter compartment, rather than movement of the glucose transporters themselves (Wang et al., 1998). One potential link between actin and known signal transduction systems is that activation of the Rho family of the small G-protein can induce a reorganization of the actin cytoskeleton (Hall, 1998; Sasaki and Takai, 1998).

It is not known whether the two astrocyte glutamate transporters GLAST and GLT-1 have distinct physiological functions. Gene disruption studies suggest that GLAST, unlike GLT-1, may not be required for neuronal survival under normal conditions in brain (Tanaka et al., 1997; Watase et al., 1998). The rapid upregulation of GLAST transport capacity in response to elevated extracellular glutamate concentrations observed here suggests that GLAST may have a unique function in feedback regulation of extracellular glutamate concentrations and glutamatergic synaptic activity. Studies by Tong and Jahr (1994) and Diamond and Jahr (1997) show that transporter binding of glutamate contributes to an early, fast reduction in synaptic glutamate concentrations in addition to the slow, later phase attributable to actual glutamate uptake [but also see Otis et al. (1996)]. Translocation of GLAST to the astrocyte cell surface could therefore affect glutamatergic transmission in two ways: by increasing the number of glutamate binding sites and by accelerating subsequent glutamate uptake.