Immunochemical Evidence for Glutamine-mediated Degradation of Glutamine Synthetase in Cultured Chinese Hamster Cells*

The specific activity of glutamine synthetase in cultured Chinese hamster cells is inversely related to the concentration of glutamine in the surrounding solution. Enzyme specific activity increases 8. to IO-fold when glutamine is removed from serum-free FlZ growth media. The induction of glutamine synthetase activity occurs only after glutamine removal and not after the removal of other amino acids (me-thionine, leucine, or isoleucine). The analyqis of the gluta-mine-mediated decrease in glutamine synthetase activity has been simplified by the finding that depression proceeds in nutrient-free buffered saline solution (141 mM NaCI, 5.4 mM KCl, and 30 rn?Z Tricine (pH 7.4) . Under these conditions, 0.1 mM cyanide blocks glutamine-mediated depression. The cyanide inhibition is reversed by the addition of 1.0 mM glucose which suggests that ATP is required for depression. Glutamine-mediated depression is temperature-dependent, occurring between 25 and 45’ with an optimum rate at 37’. Studies of the time course of induction and depression as a function of glutamine concentration suggest that glutamine regulates the rate at which the enzyme is either modified or degraded. We have employed an antibody prepared against homogeneous Chinese hamster liver glutamine synthetase to measure the amount of glutamine synthetase protein in extracts of cells containing induced or depressed levels of enzyme activity.


Glutamine
synthetase (EC 6.3.1.2) catalyzes the ATP-dependent synthesis of glutamine from glutamate and ammonia. The enzyme plays a major role in the metabolism of nitrogen because t,he amidr group of glutamine is required for the synthesis of several amino acids and nucleotides.
The complex mechanisms which cont'rol glut'amine synthetase in bacteria have been studied in detail (l-3).
Regulation includes reversible enzymatic adenylation of specific tyrosine residues and feedback inhibition by various metabolites which incorporate the amide group of glutamine.
In contrast to bacterial systems, the mechanisms by which glutamine synthetase is regulated in eucaryotes are poorly understood. Several investigators (4)(5)(6)(7)(8)(9) have reported an inverse correlation between glutamyltransferase activity in cells in tissue culture and the glutamine concentration in the supporting media. Glutamyltransferasc, measured by the capacity to form y-glutamylhydroxamate, is usually assumed to be synonymous with glutamine synt,hetase activity.
We have studied the control and regulation of glut'amine synthetase in Chinese hamster celk grown in tissue culture (10-12). Chinese hamster cells are ideally suited for biochemical and genetic studies because they can be grown easily, are pseudodiploid, and have a generation time of less than 12 hours. Using a sensitive radioisotope assay for glutamine synthetase which avoids potential problems of the y-glutamylhydroxamate assay (13, 14), we have purified and characterized the enzyme from Chinese hamster liver.
We have shown that the enzyme in Chinese hamster tissue culture cells possesses properties identical to the liver enzyme. Although glutamine sgnthetase isolated either from Chinese hamster liver or from Chinese hamster tissue culture cells is not inhibited by glutamine, the specific activity of the enzyme in tissue culture cell ext'racts is inversely related to the concentration of glutamine in the growth media. Removal of glutamine from the media results in a IO-fold increase in glutamine synthetasc activity over a 24. to 48.hour period ("induction").
The return of glutamine produces a rapid drop in en-zyme activity to its initial level ("depression" An extension of these studies is reported in this paper. The analysis has been simplified by our observation that induction of glutamine synthetase proceeds in serum-free media, and that glutamine-mediated depression can occur in nutrient-free saline solution.
The concentration of glutamine appears to regulate the rate at which glutamine synthetase is either modified or dcgraded.
This process appears to require ATP and is tem- Cell monolayers grown in media containing glutamine for 24 hours are incubated at 37' for the indicated times in serum-free media (0) or in serum-free media lacking glutamine FIG. 2 (right). Induction of glutamine synthetase by glutamine removal in the absence of essential amino acids. Cell monolayers grown in media containing glutamine for 24 hours are incubated at 37" for the indicated times in serum-free media lacking glutamine (A-A) and also lacking methionine The cell extracts in this experiment only were not assayed immediately, and the lower specific activities observed may be attributed to storage at -20" for 7 days. tions of glutamine (Fig. 3). The media are changed every 3 hours to avoid glutamine depletion and alteration of glutamine concentration.
After an initial lag, the specific activity of glutamine synthetase increases linearly at a rate independent of the final glutamine concentration. E$ect of cyanide on depression Cells grown in media lacking glutamine for 30 hours are detached with trypsin and first incubated for 60 min at 37" in suspension buffer (l), or in suspension buffer containing 0.1 mM KCN (2), 1.0 mM glucose (3), or 0.1 mM KCN and 1.0 mM glucose (4). After the first incubation either glutamine is added to a final concentration of 1.0 mM or no addition is made and the cell suspensions undergo a second incubation for 60 min at 37". The specific activity of glutamine synthetase in the cells is determined as described under "Experimental Procedures." In these experiments, induced glutamine synthetase specific activities are 0.76 to 0.82 milliunits per mg and depressed specific activities are 0.33 to 0.39 milliunits per mg. The effect of glutamine concentration on the rate of glutamine synthetase depression in cells initially containing induced levels of enzyme is determined in an analogous experiment.
Enzyme specific activity is measured at various times after replacing media lacking glutamine with serum-free media containing different concentrations of glutamine (Fig. 4). The initial rate of depression increases with increasing glutamine concentration. Regardless of whether the cells initially have induced or depressed levels of glutamine synthetase, a steady state enzyme level is eventually reached which is determined only by the glutamine concentration.
Characteristics of Glutamine-mediated Depression-The experiments with cyanide described in Table I suggest that glutaminemediated depression requires energy. Cyanide blocks the accumulation of ATP by oxidative phosphorylation and permits the dissipation of cellular ATP pools. Cells with induced levels of enzyme are incubated in suspension buffer containing cyanide, and then glutamine is added. At cyanide concentrations of 0.1 mM and above (Line 2), the addition of glutamine does not depress glutamine synthetase activity. The effect of cyanide is reversed by the addition of glucose (Line 4) which permits cells to synthesize ATP via glycolysis. Glucose by itself (Line 3) has no effect on glutamine synthetase activity either in the presence or in the absence of glutamine.
Glutamine-mediated depression is temperature-dependent (Fig. 5). Depression occurs bet.ween 25 and 45" with a maximum rate at 37". Depression does not occur at 0". Depression in the cyanide and temperature experiments is measured at 60 min after glutamine addition, and therefore the depression is less than the 8-to lo-fold observed after long term incubations in glutamine (cf. Fig. 4). Characterization of Glutamine Synihetase Antibody-The Ouchterlony double-diffusion technique demonstrates the specificity of the antigen-antibody reaction.
A single precipitin band forms when homogeneous Chinese hamster liver glutamine synthetase or a Chinese hamster liver extract is tested with the goat antiglutamine synthetase antibody (Fig. 6).
Determination After the first incubation, either glutamine is added to a final concentration of 1.0 mM or no addition is made and the cells are incubated further for 60 min at the indicated temperatures. The ratio of the enzyme activity in cells incubated without glutamine to that in cells incubated with glutamine is presented as a function of temperature.
The specific activity of glutamine synthetase in cells incubated without glutamine is 1.1 milliunits per mg from 0 to 40" and it increases 2-fold from 40 to 50" due to denaturation of other proteins. required to precipitate half the glutamine synthetase activity can be used to quantitate the amount of glutamine synthetase protein.
In Fig. 7, the unprecipitated glutamine synthetase activity is measured as a function of added anti-glutamine synthetase antibody.
Each curve represents a different initial concentration of enzyme. As the concentration of the anti-glutamine synthetase antibody increases, more glutamine synthetase precipitates and less is found in the supernatant.
Under our experimental conditions, a small amount of glutamine synthetase activity (approximately 0.1 to 0.2 nmol of glutamine per 30 min) remains unprecipitated even at high anti-glutamine synthetase antibody concentrations.
This unprecipitated enzyme may be attributed to incomplete precipitation of the enzyme-goat antibody complex by the rabbit anti-goat y-globulin antibody.
Apparently, in our assay conditions the anti-goat y-globulin antibody is not in large excess. Ten-fold higher amounts of anti- The amount of glutamine synthetase protein corresponding to the initial activity is based on the known specific activity of homogeneous glutamine synthetase, 9.5 units per mg (10). goat y-globulin antibody than used in these experiments reduces the unprecipitable glutamine synthetase 2-to 3-fold. Nevertheless, the data in Fig. 8 indicate that the assay is satisfactory, and the routine use of higher levels of rabbit anti-goat y-globulin antibody was impractica1.
The data in Fig. 7A are obtained with glutamine synthetase from extracts of Chinese hamster cells having induced levels of enzyme activity.
Analogous curves are obtained using homogeneous Chinese hamster liver glutamine synthetase (Fig. 7B).
The concentration of antibody for 50% precipitation is defined as that amount at which the enzyme activity remaining in the supernatant is the average of the initial and final levels.

Glutamine synthetase specific activity in cultured Chinese hamster cells is inversely related
to the concentrat,ion of glutamine in cell media. Our attempts to demonstrate and to characterize glutamine-mediated depression of enzyme activity in cell-free extracts have not been successful. We have sought to better characterize enzyme regulation in viva as a prelude to more systematic st'udies in vitro.
The current work demonstrates that both induction by glutamine removal and depression by glutamine addition occur in the absence of serum. Hence, glutamine-mediated regulation can be studied in defined media. Furthermore, our observation that depression can occur in a saline solution has enabled us to examine potential factors affecting glutamine depression.
Cell suspensions rather than monolayers are employed in these latter studies to facilitate thorough washing of cells and removal of glutamine.
The slower rates of depression observed under these conditions compared to those found in previous studies (11) may be due to the absence of media components which affect the rate but not the extent of depression. Induction of glutamine synthetase results from glutamine removal specifically and is not caused by general amino acid deprivation.
Starvation for methionine, leucine, or isoleucine does not increase enzyme activity when glutamine is present. Moreover, removal of any of these three essential amino acids does not prevent partial induction when glutamine is simultaneously removed.
We previously have shown that cycloheximide blocks induction implying that protein synthesis is required for increase in enzyme activity (11). The removal of an essential amino acid  Fig. 8. Data corresponding to the curves in Fig. 9 (A, 0).
Data expected if depression of glutamine synthetase resulted from a glutamine-mediated modification (A, 0). These latter points are calculated from the ratios of induced (2.6 milliunits per mg) and depressed pecific activities assuming that the calibration curve is linear 07 :r a larger range than indicated in Fig. 8.
The amount of anti-gluta.mine synthetase antibody required to precipitate 50% of the glutamine synthetase activity is a linear function of the initial enzyme concentration (Fig. 8). Data for Fig. 8 come from immunoprecipitation reactions both with extracts from Chinese hamster cells having induced levels of glutamine syrkhetase activity and from reactions with homogeneous Chinese hamster liver glutamine synthetase.
Since both sets of data fall on the same line, glutamine synthetase in extracts of Chinese hamster cells with induced levels of activity must have the same relative amounts of activity and antigenicity as those in Chinese hamster liver.
We can define the molecular specific activit'y as the ratio of activity to antigenicity.
The molecular specific activity is a measure of the activity of an enzyme molecule.
The data in Fig. 8 suggest that the immunoprecipitation reaction can be used to compare the molecular specific activity of glutamine synthetase in extracts of cells with induced levels of enzyme activity with that in extracts of cells with depressed levels of enzyme activity.
Measurement of Glutamine Synthetase Protein in Cell Extracts Containing Depressed Levels of Glutamine Synthetase Activity-The data in Fig. 9 illustrate t,he results of an immunoprecipitation experiment with extract,s of cells which have been grown in the presence of glutamine for short (2 hours) and long (24 hours) periods of time.
Both extracts require somewhat more antiglutamine synthetase antibody for precipitation than do extracts containing induced levels of enzyme at the same initial activity. The calibration curve (Fig. 8) is redrawn in the inset in Fig. 9 along with the data obtained from the immunoprecipitation experiments with the extracts of cells containing depressed levels of enzyme. Also shown are the data that would be expected if glutamine-mediated depression resulted from a modification of glutamine synthetase to a less catalytic but antigenically identical form.
(e.g. methionine) decreases the rate of incorporation of another essential amino acid (e.g. [3H]leucine) by approximately 50% (data not shown), indicating that protein synthesis continues at a decreased rate during essential amino acid starvation.
Since pools of methionine and leucine are low in Chinese hamster cells (data not shown), the necessary amino acids for synthesis of glutamine synthetase during essential amino acid starvation are probably provided by protein turnover.
The increase in glutamine synthetase specific activity is not due to selected cell survival or accelerated degradation of non-glutamine synthetase protein since viable cell survival is 100% and total protein before and after induction is nearly identical.
Glutamine-mediated regulation of glutamine synthetase specific activity could operate by affecting either the rate of enzyme synthesis or the rate of enzyme degradation.
During induction, enzyme specific activity linearly increases after an initial lag at a rate which is independent of the glutamine concentration. In contrast, the addition of glutamine to cells with induced activity results in an exponential decrease in activity at an initial rate which is dependent upon glutamine concentration.
Similar results have been obtained by Kulka and Cohen (9) in steroidtreated hepatoma tissue culture cells. We conclude that glutamine either directly or indirectly regulates the rate of depression of glutamine synthet'ase, and does not affect the rate of enzyme synthesis.
Depressiou could occur either by degradation or by modification to produce an enzyme with a lower molecular specific activity.
Stadtman and co-workers have found that covalent adenylylation (promoted by glutamine) controls the activity of Escherichia co& glutamine synthetase (2). The adenylylated enzyme possesses lower specific activity and displays different susceptibilities to inhibition by metal ions and cellular metabolites. Although induced and depressed Chinese hamster cell glutamine synthetase appear to be identical by several catalytic and physical criteria (12), t,he existence of two enzyme species differing in their molecular specific activities was not precluded.
Regulation by specific drgradat#ion or inactivation by a variety of mechanisms is well documented in eucaryotic cells (16)(17)(18)(19)(20). The speed and apparent specificity in the response of glutamine synthetase to glutamine are similar to class-specific proteolysis now implicated in enzyme regulation by Katunuma,Holzer,.
Conformation appears to be an important factor in determining the susceptibility of protein structure to cleavage and degradation.
Proteins undergo constant fluctuations in their three-dimensional structures (26,27). As some conformat,ions are more labile t,han others, the relative stabilization of a particular conformation by binding some component or by covalent modification could modulate the half-life of the molecule (24,25,28,29).
Our studies reIating glutamine synthetase activity and ATP (10-12) may also reflect the involvement of conformat,ional equilibria in enzyme regulation.
The experiments with cyanide reported in this paper indicate that ATP may be essential for gluxamine-mediated depression of glutamine synthetase.
Thus, in the presence of ATP, glutamine or some component proportional to glutamine concentration could bind to glutamine synthetase and enhance its susceptibility to degradation.
Immunoprecipitation of specific enzyme protein can be used to determine if a change in specific activity of an enzyme results from a corresponding change in level of the enzyme protein (30). The immunoprecipitation method most often used involves the addition of increasing amounts of enzyme activity to a constant amount of antibody.
The antibody-antigen complex is removed by centrifugation and the residual activity left in the supernatant is measured.
We find that this immunoprecipitation titration technique is not satisfactorily reproducible at the low glutamine synthetase concentrations present in Chinese hamster cell extracts. The concentration of glut'amine synthetase in extracts of Chinese hamster cells with induced levels of enzyme is lo-to 100.fold less than t,hat employed in other immunoprecipitation studies (30)(31)(32)(33)(34)(35)(36)(37)(38)(39).
Moreover, the concentration of glutamine synthetase in extracts of cells with depressed levels of enzyme is IO-fold lower.
When the amount of enzyme is kept constant and the amount of antibody is varied, a linear calibration curve can be obtained (Fig. 8) which relates the amount of glutamine syntketase protein to the amount of antibody needed to precipitate half the enzyme activity.
As little as 2 ng of glutamine synthetase protein can be detected.
The data in Fig. 8 show that homogeneous glutamine synthetase from Chinese hamster liver and glutamine synthetase in cell extracts containing induced levels of enzyme display identical immunoprecipitation reactions. Thus, glutamine synthetase in cell extracts containing induced levels of enzyme possesses the same molecular specific activity as homogeneous liver glutamine synthetase.
Glutamine synthetase in extracts of cells grown for either short (2 hours) or long (24 hours) times in the presence of glutamine and which therefore contain depressed levels of activity have only a slightly lower molecular specific activity.
If a glutamine-dependent modification reaction were responsible for the 8-to lo-fold decrease in glutamine synthetase activity, one Protein ) q +G Glutamine synthetase is synthesized at a rate independent of the glutamine concentration.
Glutamine or some component directly proportional to the glutamine concentration (G) reversibly binds to the enzyme making it more susceptible to degradation.
would expect that' the molecular specific activity of glutamine synthetase in extracts of cells containing depressed levels of activity would be 8-to IO-fold less than that for glutamine synthetase in extracts of cells containing induced levels of enzyme (Fig. 9, inset). The fact that we see only a slight decrease in molecular specific activity suggests that most of the glutamine synt,hetase protein that is present in extracts of cells containing induced levels of enzyme is either degraded upon depression of the enzyme activity or drast,ically altered in antigenicity.
The latter possibility seems unlikely since a protein as large as glutamine synthetase (10) probably has many antigenic sites, only a few of which are likely to be altered by a modification.
A model for glutaminc-mediated regulation of Chinese hamster glut,amine synthetase which fit's our data is illustrated in Fig. 10. Glutamine synthetase enzyme (E) is assumed to be synthesized at a rate independent of the glutamine concentration. Glutamine or some component directly proport,ional to the glutamine concentration (G) binds reversibly to the enzyme, making it more susceptible to degradation.
The model predict#s t#hat the molecular specific activities of glutamine synthetase in cultured Chinese hamster cells grown for long times either in the presence or in the absence of glutamine should be the same. If the degradation step is fast, the amount of glut'amine synthet,ase protein will decrease at the same rate as the decrease in activity.
Under these conditions, the molecular specific activity also will be nearly the same in cells depressed for short times. Alt'hough we find that the molecular specific activity in cells depressed for short and long times is slightly less than that in cells containing induced levels of enzyme, this difference could result from the presence of a small amount of partly degraded but ant'igenic enzyme molecules.
A mathematical analysis of our model is consistent with the steady state levels of glutamine synthetasc observed at different glutamine concent,rations (Figs. 3 and 4). The rate of enzyme degradation is given by where kl is a rat'e constant, k, is an affinity constant, and E is the relative fraction of cell protein that is glutamine synthetase. If EI is the maximum relative fract,ion of enzyme protein in cells with induced levels of enzyme activity growing in the absence of glutamine, then: E = EI -(kak;iy;;n,) where SI is the specific activity of glutamine synthetase in extracts of cells with induced levels of activity growing in the absence of glutamine. This equation can be rearranged to give: (kS1)(Sr -S)+ = k,[Gln]-1 + (kl + 1). As predicted by this model, a plot of (S1 -S)-l versus [Gin]-l for the steady state specific activities from Figs. 3 and 4 gives a straight line (Fig. 11).
In conclusion, the data present'ed in this paper suggest that the glutamine-mediated depression of glutamine synthetase activity results from degradation of enzyme molecules. Degradabion might occur by a glutamine-activated enzyme-specific protease or by a glutamine-mediated alteration in the enzyme's structure enhancing its susceptibility to degradation. Unfortunately, the molecular components required for depression appear to be inactivated when cells are lysed. Glutamine synthetase is remarkably stable in crude cell extracts in. vitro in the presence or absence of glutamine.
In future studies, we l~ill attempt to identify and characterize the molecular mechanisms responsible for glutamine-mediated degradation of glutamine synthetase.