Induction of high affinity glutamate transport activity by amino acid deprivation in renal epithelial cells does not involve an increase in the amount of transporter protein.

In renal epithelial cells amino acid deprivation induces an increase in L-Asp transport with a doubling of the V and no change in K (4.5 μM) in a cycloheximide-sensitive process. The induction of sodium-dependent L-aspartate transport was inhibited by single amino acids that are metabolized to produce glutamate but not by those that do not produce glutamate. The transaminase inhibitor aminooxyacetate in glutamine-free medium caused a decrease in cell glutamate content and an induction of glutamate transport. In complete medium aminooxyacetate neither decreased cell glutamate nor increased transport activity. These results are consistent with a triggering of induction of transport by low intracellular glutamate concentrations. High affinity glutamate transport in these cells is mediated by the excitatory amino acid carrier 1 (EAAC1) gene product. Western blotting using antibodies to the C-terminal region of EAAC1 showed that there is no increase in the amount of EAAC1 protein on prolonged incubation in amino acid-free medium. Conversely, the induction of high affinity glutamate transport by hyperosmotic shock was accompanied by an increase in EAAC1 protein. It is proposed that low glutamate levels lead to the induction of a putative protein that activates the EAAC1 transporter. A model illustrating such a mechanism is described.

In renal epithelial cells amino acid deprivation induces an increase in L-Asp transport with a doubling of the V max and no change in K m (4.5 M) in a cycloheximide-sensitive process. The induction of sodium-dependent L-aspartate transport was inhibited by single amino acids that are metabolized to produce glutamate but not by those that do not produce glutamate. The transaminase inhibitor aminooxyacetate in glutamine-free medium caused a decrease in cell glutamate content and an induction of glutamate transport. In complete medium aminooxyacetate neither decreased cell glutamate nor increased transport activity. These results are consistent with a triggering of induction of transport by low intracellular glutamate concentrations. High affinity glutamate transport in these cells is mediated by the excitatory amino acid carrier 1 (EAAC1) gene product. Western blotting using antibodies to the C-terminal region of EAAC1 showed that there is no increase in the amount of EAAC1 protein on prolonged incubation in amino acid-free medium. Conversely, the induction of high affinity glutamate transport by hyperosmotic shock was accompanied by an increase in EAAC1 protein. It is proposed that low glutamate levels lead to the induction of a putative protein that activates the EAAC1 transporter. A model illustrating such a mechanism is described.
In mammalian cells intracellular glutamate concentrations are maintained at a high level by the presence of active transport systems for glutamate in the plasma membrane. A number of different glutamate transporters have been kinetically characterized in various cell types including the Na ϩ -independent transporter System x c Ϫ and the high affinity (K m 2-50 M) sodium-dependent glutamate transporter X AG Ϫ. System X AG Ϫ (1), which is the major glutamate transporter in most nonneuronal cell types, is characterized by its specificity for L-glutamate but nonselectivity for aspartate as both the D-and Lisomers are transported. The transport process is electrogenic, and it is suggested that two sodium ions and one glutamate are transported into the cell in exchange for one potassium and one bicarbonate ion (2).
The renal bovine epithelial cell line (NBL-1) has been used in our laboratory as a model system to study the regulation of amino acid transport (10 -12). These cells, which are probably of distal tubule origin, express high activities of Na ϩ -dependent glutamate transport which have properties similar to System X AG Ϫ as studied in other cell types (1). In these cells glutamate transport is highly regulated. When NBL-1 cells are starved of amino acids, there is an increase in X AG Ϫ transport activity with a doubling of V max and no change in the K m after 10 h (13). Incubation of NBL-1 cells in 200 mM sucrose also induced glutamate transport with a 3-fold increase in V max and no change in K m (14). System X AG Ϫ in NBL-1 cells has the same kinetic properties as glutamate transport induced in Xenopus oocytes by injection of EAAC1 cRNA; since EAAC1 is known to be expressed in kidney while the other transporters are mainly restricted to brain, it is likely that EAAC1 encodes the NBL-1 cell glutamate transporter. We have shown that induction of System X AG Ϫ activity by hyperosmotic shock in NBL-1 cells is accompanied by a 3-fold increase in EAAC1-specific mRNA (14). Conversely, there is no such increase in EAAC1 mRNA during induction of transport activity by amino acid deprivation (13).
In this paper the induction of System X AG Ϫ by amino acid deprivation is further characterized, and sequence information is exploited to raise an antibody to show changes in the protein level during the induction of transport. We have shown that the internal glutamate level is the major determinant of the induction of System X AG Ϫ; furthermore, there is no change in the amount of transport protein during the increase in transport activity.
were synthesized in the Department of Biochemistry. Tissue culture reagents were purchased from Life Technologies, Inc. (Paisley, UK) except the dialyzed newborn calf serum which was from Sigma.
Cell Culture-Unless indicated otherwise the NBL-1 cells were seeded at a density of 6 ϫ 10 3 cells/ml into 35-mm Petri dishes in Ham's F-12 medium supplemented with 10% (v/v) fetal bovine serum, 2 mM glutamine, and antibiotics as described previously (11). The cells were fed every other day until used in experiments, typically after 4 days. The amino acid-free medium used contained the inorganic salts of Ham's F-12 medium together with 10 mM glucose and 0.1% bovine serum albumin (11). The glutamine-free medium contained all the ingredients of Ham's F-12 except glutamine and was supplemented with 10% (v/v) dialyzed newborn calf serum.
Transport Measurements-System X AG Ϫ activity was measured as the initial rate of sodium-dependent uptake of 50 M [U-14 C]L-aspartate into NBL-1 cells over 5 min at 20°C in the presence of 0.5 mM aminooxyacetic acid to inhibit aspartate metabolism (13). In this system Na ϩ -dependent L-aspartate transport is linear with time over 10 min (13). Sodium-independent transport was measured in medium where NaCl was replaced by an equal concentration of choline chloride. At 50 M L-Asp the Na ϩ -independent rate was less than 15% of the sodiumdependent rate and did not change under any of the conditions used. Cell protein was measured as described (15).
Anti-peptide Antibody-The anti-peptide antibody was raised against a synthetic C-terminal peptide CAVDKSDTISFTQTSQF (amino acids 509 -524 of the EAAC1 sequence (3) with an cysteine added to the N-terminal end). The peptide was linked to keyhole limpet hemocyanin via the terminal cysteine using sulfo-SMCC. The antibody was purified by passing down a column of the synthetic peptide cross-linked to CNBr-activated Sepharose 4B beads.
Cloning of the Hydrophilic Loop Region of EAAC1 and Preparation of an Anti-fusion Protein Antibody-The cDNA encoding the putative extracellular loop between helices 3 and 4 (amino acids 117-215) was first amplified from NBL-1 cell cDNA using nested PCR. The flanking primers were 5Ј-GCT/C TTT/C CCC GGA/C GAG ATT/C CTG/T/C ATG and 5Ј-TCA/G TAA/G/C AGA/G GCG GTT/G CCA/G TCC AT.
The product of the first reaction was used as template for second round PCR. The nested primers were 5Ј-T TTT GGA TCC AAG CCC/T GGC/A/G GTG/C ACC CAG/A AAG and 5Ј-TT TTG GAT CCG CCC/G/A AGC/G ACA/G TTG/A ATG/T CCG/A TC.
The PCR product was sequenced and shown to have 83.75% identity with the rabbit intestine EAAC1 cDNA sequence and 78.5% identity at the protein level. Because problems were encountered in direct ligation into the prokaryotic expression vector pGEX2T, the DNA was bluntended and cloned into EcoRV-cut pBluescript, which had been pretreated with calf intestinal alkaline phosphatase. The construct was transformed into Ca 2ϩ -competent Escherichia coli XL-1 by heat shock. Using the BamHI sites engineered into the primers, the insert was excised from pBluescript and ligated into pGEX2T, which was used to transform XL-1 as before. The construct was sequenced across the insert site and shown not to have any frameshift mutations. Once the absorption of the transformed cells at 600 nm reached 0.3, glutathione-S-transferase-loop construct expression was induced for 1 1 ⁄2 h with 1 mM isopropyl-1-thio-␤-D-galactopyranoside. 10 out of 20 transformants expressed the glutathione-S-transferase-loop construct in the correct orientation. The bacterial proteins were separated on SDS-PAGE, and the fusion protein was electroeluted overnight at 50 mA.
Western Blots-Cell extracts were prepared as follows. Cell cultures were washed twice in ice-cold phosphate-buffered saline (PBS). The cells were scraped off into ice-cold PBS in the presence of protease inhibitors (2 mM benzamidine, 1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, and PAL (1 g/ml each of pepstatin, antipain, and leupeptin)) and centrifuged. The PBS was removed, and the cells were resuspended in 20 mM Tris-HCl, pH 7.4, before dissolving in the neutral detergent MEGA-10 to a final concentration of 1% in the presence of protease inhibitors. 25 g of the samples were then separated on SDS-PAGE under reducing conditions. The proteins were transferred in the absence of SDS using a semi-dry blotting apparatus, 1 h, 10 V.
The blots were blocked for 20 min in PBS/0.02% Tween containing 5% dried milk powder. The antipeptide antibody was added for 1 h in the same buffer, and the blots were then washed three times in PBS/ 0.02% Tween before anti-rabbit horseradish peroxidase conjugate was added at 1 l/ml in 5% powdered milk/PBS/0.02% Tween for 45 min. The blots were finally washed twice in PBS/0.02% Tween and then once in PBS before development by enhanced chemiluminescence.
Measurement of Internal Glutamate Concentration-Glutamate in neutralized acid extracts of whole cells was measured by the method of Bernt and Bergmeyer (16).
Sequencing-Sequencing of the DNA cloned into pBluescript and pGEX2T was carried out using 35 S-dATP and the Sequenase sequencing kit according to the manufacturer's instructions. The KS and T7 primers were used for sequencing pBluescript. A specific primer GCAT-GGCCTTTGCAGGG was used for sequencing across the pGEX2T multiple cloning site.

Characterization of the Induction of System X AG Ϫ-Activity by
Amino Acid Deprivation in NBL-1 Cells-Total amino acid deprivation causes an increase in the V max of Na ϩ -dependent L-aspartate transport in NBL-1 cells. This process is sensitive to cycloheximide and is maximal after 10 h (13). On repeating this work it was found that the magnitude of the stimulation depended on the degree of confluency of the cells (Fig. 1). There was a 300% induction by amino acid starvation 3 days after seeding the cells (confluency approximately 80%), but the degree of induction decreased as the cells became more confluent and was relatively low in confluent cells. Subsequent experiments were performed under conditions where the induction was maximal, as indicated in the figure legends.
Addition of certain single amino acids to the amino acid-free medium reduced the induction of System X AG Ϫ activity, and this is further characterized in Table I. All the substrates of System X AG Ϫ when added at 1 mM concentration at zero time caused 50% inhibition when transport was measured 24 h later. In addition to the substrates of System X, the amino acids L-Asn, L-Ala, and L-Gln, which can be converted to glutamate by transamination, reduced the induction of transport activity. Conversely, D-Gln, L-Cys, and L-Leu, which cannot produce intracellular glutamate in these cells, did not influence the induction of transport. D-Glu, which is a poor substrate for System X AG Ϫ, also inhibited the induction in accordance with previous work (13).
In order to confirm that the effect of added amino acids on the apparent induction of transport was not in fact due to trans-inhibition of uptake by differential internal concentrations of glutamate, cells were incubated with various single amino acids (1 mM) in amino acid-free medium for 1 h allowing intracellular glutamate accumulation but not allowing enough time for protein synthesis to occur. In all cases there was no increase above the rate in amino acid-free medium (not shown). Three days after seeding the NBL-1 cells, the culture medium was either changed to fresh normal medium or the cells were washed once with amino acid-free medium and cultured in the same amino acid-free medium. The degree of confluency was assessed at this stage, and the transport experiments were performed 24 h later. This was repeated on the two subsequent days. The initial rate of Na ϩ -dependent aspartate transport is shown as the mean Ϯ S.E. of values obtained from three Petri dishes in each case.
It was previously shown, and is confirmed here, that the induction of transport activity by amino acid-free medium was completely inhibited by cycloheximide, confirming that the effect was due to protein synthesis.
The results in Table I suggest that the intracellular glutamate concentration may determine the level of expression of System X AG Ϫ activity. An independent way of manipulating intracellular glutamate levels is by the use of aminooxyacetic acid (AOA), which is a specific inhibitor of transaminases. AOA prevents the formation of glutamate from other amino acids. Fig. 2 shows an experiment in which cells were incubated in the presence and absence of AOA in either normal medium or glutamine-free medium. Transport was measured after 24 h. AOA caused an induction of transport when present in the glutamine-free medium but not in normal medium. In the absence of AOA, omission of glutamine produced only a small induction. In the absence of glutamine and the presence of AOA glutamate can be produced neither by glutamine hydrolysis nor by transamination of other amino acids, and intracellular glutamate levels would be predicted to fall. The fact that AOA did not induce transport activity in the presence of glutamine indicates that AOA per se was not responsible for the induction. Under these conditions intracellular glutamate is predicted to be high since glutamate produced by glutaminase cannot be transaminated. Since the rates of Na ϩ -dependent aspartate transport were not markedly reduced by AOA under any conditions, it is unlikely that AOA was significantly damaging cell function.
In order to confirm this interpretation, the cell glutamate content was determined enzymatically. Fig. 3 shows that on switching cells to amino acid-free medium the cellular glutamate content fell from 30 nmol/mg protein to 13 nmol/mg protein over a period of 5 h. In cells transferred to fresh normal medium at zero time, the cellular glutamate concentration fell much more slowly. Under the conditions corresponding to the experiment shown in Fig. 2, the internal glutamate contents at 1 h were determined. In glutamine-free media ϩ 0.5 mM AOA the cell glutamate content fell to 15.0 Ϯ 0.6 nmol/mg protein, while in normal media in the presence of 0.5 mM AOA the cell glutamate content was 29.2 Ϯ 1.0 nmol/mg protein, which is similar to the control values. These results suggest that the induction of System X AG Ϫ is triggered when the glutamate level falls to about 15 nmol/mg protein. This occurs within an hour in the presence of AOA and the absence of glutamine or in amino acid-free cells after 2-3 h. Glutamate levels did not reach this value in cells cultured in normal medium.
Induction of System X AG Ϫ-Activity by Tunicamycin-Tuni-camycin is an inhibitor of protein glycosylation. Incubation of cells with 0.1 g/ml tunicamycin in amino acid-free medium for 24 h apparently reduced the amino acid starvation-dependent induction of aspartate transport, suggesting that glycosylation of a protein is a necessary step in this process (Fig. 4). However, the addition of 0.1 g/ml tunicamycin to cells in normal medium for 24 h itself induced transport activity. The induction was cycloheximide-sensitive (Fig. 4). This effect was characterized by an increase in the V max from 110 to 180 pmol/mg/min whereas the K m for L-aspartate was unchanged at 4.6 M (results not shown). The half-maximal effect was at 0.055 g/ml tunicamycin, which is a concentration at which the effect of this compound on protein glycosylation should be specific. Tunica-  (1 mM) in the culture medium on the induction of L-Asp transport Three days after seeding, NBL-1 cells were transferred to fresh media as indicated below. The initial rate of Na ϩ -dependent L-asp transport was determined 24 h later. The data presented are means Ϯ S.E. of three determinations in each case.

Medium
Rate of L-asp transport

pmol/mg/min
Complete medium (Hams' F-12) 87.8 Ϯ 0.9 a Amino acid-free medium 238 Ϯ 4 Amino acid-free medium ϩ 1 mM L-Asp 164 Ϯ 4 a Amino acid-free medium ϩ 1 mM L-Glu 163 Ϯ 3 a Amino acid-free medium ϩ 1 mM D-Glu 159 Ϯ 2 a Amino acid-free medium ϩ 1 mM L-Gln 156 Ϯ 3 a Amino acid-free medium ϩ 1 mM D-Gln 210 Ϯ 5 Amino acid-free medium ϩ 1 mM L-Asn 146 Ϯ 4 a Amino acid-free medium ϩ 1 mM L-Ala 157 Ϯ 2 a Amino acid-free medium ϩ 1 mM L-Cys 207 Ϯ 4 Amino acid-free medium ϩ 1 mM L-Leu 211 Ϯ 6 a p Ͻ 0.005 versus rate in amino acid-free medium. FIG. 2. Effect of aminooxyacetic acid in the presence and absence of glutamine on the induction of L-aspartate transport activity. Three days after seeding the cells the medium was changed to either normal medium or glutamine-free medium in the presence or absence of 0.5 mM aminooxyacetic acid as indicated in the bar chart. Before incubation in glutamine-free medium, the cells were washed once with glutamine-free medium to remove any residual glutamine. 24 h after this, the initial rate of transport of L-aspartate was measured. Data presented are mean Ϯ S.E. of three determinations in each case.

FIG. 3. Time course of changes in cellular L-glutamate content.
NBL-1 cells were seeded in T75 flasks at 3 ϫ 10 3 cells/ml retaining the same cell density/cm 2 as the seeding in the 35-mm Petri dishes. Three days after seeding, the cells were incubated in normal medium or washed once in amino acid-free medium before incubation in amino acid-free medium for the times indicated. The glutamate contents in cells incubated in normal medium are shown as single points, and the glutamate contents of cells incubated in amino acid-free medium are the mean Ϯ S.E. of values from three separate flasks of cells in each case. mycin is known to exert a stress effect on cells that is characterized by an increase in the 78-kDa glucose-regulated protein (GRP78) mRNA (17).
Changes in EAAC1 Protein Levels During Induction of Activity by Amino Acid Deprivation, Hyperosmotic Stress, and Tunicamycin-The kinetic parameters and substrate specificity of System X AG Ϫ activity in NBL-1 cells (13) are almost identical to those of the EAAC1 gene product expressed in Xenopus oocytes (3). Primers designed to two internal EAAC1 sequences from rabbit intestine produced PCR products of the expected size using NBL-1 cell cDNA as a template; partial sequences of the PCR products were in close agreement with the published EAAC1 sequence (3). Furthermore, induction of System X AG Ϫ activity by exposure of NBL-1 cells to hyperosmotic media was accompanied by a 3-fold increase in EAAC1 mRNA levels (14). EAAC1 is the only high affinity glutamate transporter known to be present in the kidney (3,6). These considerations, taken together, show that the EAAC1 gene product is responsible for high affinity glutamate transport in NBL-1 cells. In order to measure changes in the level of the EAAC1 protein during amino acid deprivation, two specific antibodies to the protein were produced. One antibody was raised to the C-terminal peptide of EAAC1 whereas the other was raised to a glutathione-S-transferase fusion protein containing the 100-amino acid hydrophilic loop between amino acids 117 and 214.
The purified anti-peptide C-terminal antibody recognized a major band at 64 kDa in NBL-1 cells, together with some minor bands (Fig. 5). This is reasonably consistent with the molecular mass of the EAAC1 protein from rabbit intestine which is predicted to be 57 kDa. The molecule contains four potential glycosylation sites, but the extent of glycosylation is not known. Fig. 5 also shows that the anti-loop antibody recognized a major band of the same size as that recognized by the C-terminal anti-peptide antibody. The fact that antibodies raised to different regions of the EAAC1 protein recognize the same protein is good evidence that this protein is the EAAC1 gene product. Western blots (not shown) indicate that the C-terminal antibody recognized a protein of the same molecular weight in rat brain. This is also consistent with the molecular weight of the protein recognized in rat brain by another antibody raised to the C-terminal of EAAC1 (7). Fig. 6 shows Western blots of NBL-1 cells incubated for 24 h in normal medium, amino acid-free medium, normal medium ϩ 200 mM sucrose, and normal medium ϩ 0.1 g/ml tunicamycin. No change in the protein level was observed between normal and amino acid-starved cells, although an increase in transport under these conditions has been shown to occur. However, cells exposed to sucrose or tunicamycin (other conditions that induce transport activity) showed a very significant increase in the amount of protein detected by the anti-peptide antibody. Table  II quantifies the changes in protein over a number of different experiments.
Table II also reports the results of a similar experiment where transport was induced by the addition of 0.5 mM AOA and in the absence of 1 mM L-Gln. There was no change in the amount of protein even though an increase in L-Asp transport occurred under these conditions (see Fig. 2).

DISCUSSION
The results presented above indicate that the necessary condition for the induction of glutamate transport by amino acid deprivation in NBL-1 cells is the presence of a cellular glutamate content of less than about 15 nmol/mg. Since the intracellular volume in these cells is about 6.2 l/mg protein (10), this would correspond to an intracellular concentration of 2.5 mM. When glutamate is depleted from 30 to about 15 nmol/mg the transport activity starts increasing after a period of about 5-6 h. These conclusions follow from the following facts. (i) Incubation of cells in amino acid-free medium leads to a reduction of glutamate levels to 15 nmol/mg within 2-3 h. (ii) Single amino acids that can be metabolized to glutamate prevent induction whereas those that cannot be metabolized cannot prevent induction (Table I) 1) and anti-fusion protein (lane 2) antibody recognize a band of the same size. Three days after seeding, the medium on NBL-1 cells was changed to fresh complete medium. 24 h later whole NBL-1 cell extracts were prepared as described under "Experimental Procedures," separated by SDS-PAGE, and transferred to nitrocellulose. Duplicate 25-g samples were probed with either the antipeptide antibody or anti-fusion protein antibody.  /ml tunicamycin (lane 4). Three days after seeding, the medium was changed to that indicated. 24 h later whole cell extracts were prepared as described under "Experimental Procedures," and 25-g samples were separated on SDS-PAGE, transferred to nitrocellulose, and probed with the anti-peptide EAAC1 antibody. of glutamine reduces glutamate levels to 15 nmol/mg and causes the induction of transport. (iv) AOA in the presence of glutamine neither reduces glutamate levels nor induces transport. This is likely to be a physiologically important mechanism for maintaining glutamate levels.
Western blots have shown that there is no change in the EAAC1 protein level during amino acid deprivation even though there is an increase in transport. Under other conditions where transport has been shown to be induced, i.e. exposure to hyperosmotic medium or tunicamycin, a clear increase in the amount of protein detected by the C-terminal antipeptide antibody is observed. The increase in transport activity is dependent on protein synthesis and is not due to differential transinhibition as a result of different internal glutamate concentrations. Also it has been previously shown in this laboratory that EAAC1 mRNA levels do not increase during amino acid deprivation, although these mRNA levels do increase as a result of hyperosmotic shock (14) and tunicamycin treatment. 2 Since the induction of transport activity requires protein synthesis but the amount of EAAC1 protein itself does not increase, we postulate that the induction of a putative EAAC1activating protein is responsible for the increase in the rate of aspartate transport. As the K m is unchanged on incubating the cells in amino acid-free medium (13), it is unlikely that this is due to the induction of a different glutamate transporter. Since the activation is reduced by tunicamycin, this suggests the putative activating protein may well be a glycoprotein.
Tunicamycin itself in normal medium caused increases in EAAC1 mRNA and protein. The mechanism of this effect is not clear but is likely to be related to a stress effect possibly triggered by the presence of malfolded proteins as has been suggested for GRP78 (18). As there is a 4-fold increase in EAAC1 protein, but only a doubling in the rate of transport, it appears that some of EAAC1 protein is not reaching the plasma membrane due to mistargetting in the absence of glycosylation.
The question arises as to how the cell is able to detect changes in the cellular glutamate level in the range of 30 to 15 nmol/mg. There are a number of proteins that regulate amino acid transport in bacteria such as the leucine-responsive regulatory protein in E. coli (19) and the glutamate uptake regulatory protein in Zymomonas mobilis. This protein has a helixturn-helix motif that is typical of a transcription factor and has been shown by gel retardation assays to bind the regulatory region of the E. coli gene gltP, which encodes a proton symporter for glutamate and aspartate (20). The results in this paper are consistent with the presence of a low affinity glutamate binding protein (GBP) that can act in one of two ways to switch on the synthesis of a protein that activates System X AG Ϫ (Fig. 7). One possibility is that GBP is a transcription factor that binds L-glutamate with low affinity. Alternatively, the GBP could be an mRNA-stabilizing factor or an mRNA-binding protein affecting mRNA translation. In either case GBP-glutamate complex is assumed to be inactive. Relatively more of the active form of the protein would be present at low glutamate concentrations. Since D-glutamate also inhibits the amino acid deprivation-induced increase in System X AG Ϫ activity, GBP must also be assumed to bind D-glutamate.
In NBL-1 cells System A is also induced by amino acid deprivation in a process that is protein synthesis-dependent, sensitive to tunicamycin, and reversed or prevented by the addition of single amino acids (11); these results are consistent with earlier work on hepatocytes (21). Since System A has not yet been cloned, no definitive mechanism for this effect has been established. There are indications that the induction of System A involves the synthesis of a hypothetical transport activating protein (21) rather than the System A transport protein itself. This conclusion has been reinforced by studies of Chinese hamster ovary cell mutants that do not induce System A activity on amino acid deprivation (Ref. 22; for review see Ref. 23). It is possible that amino acid deprivation leads to the 2 B. Nicholson, unpublished results.

TABLE II Quantification of the relative amount of EAAC1 protein in NBL-1 cells incubated for 24 h in various conditions
Three days after seeding the medium was changed to that indicated in the table. 24 hours later the whole cell extracts were prepared, and 25-g samples were probed with the anti-peptide antibody as described in Fig. 5. The area under the 64-kDa band was determined using a gel scanner. For the cells in complete medium this value was normalized to 1. The data were obtained from the number of separate tracks shown, each track representing cells from a different Petri dish.

Medium
Relative  7. A model to account for the induction of System X AG ؊ activity by low intracellular glutamate concentrations. The existence of a GBP is proposed. The free form of this protein causes an increase in the rate of synthesis of a putative System X AG Ϫ-activating protein, whereas the glutamate-bound form is inactive. There are two possible mechanisms by which the free GBP could act, either by interacting with a promoter for the activator protein or by interaction with the mRNA for the activator protein. Points of interaction of inhibitors are shown. Induction of System X AG Ϫ transport activity by exposure of cells to hyperosmotic medium is mediated by a transcriptionally regulated increase in the EAAC 1 protein itself. synthesis of one or more glycoproteins that act as activators of amino acid transport proteins in cell membranes, thus assisting in maintenance of the intracellular amino acid pool under these conditions.