Coordinated Regulation of Glutamine:Fructose-6-phosphate Amidotransferase Activity by Insulin, Glucose, and Glutamine ROLE OF HEXOSAMINE BIOSYNTHESIS IN ENZYME REGULATION*

We reported previously that glutamine:F-6-P ami- dotransferase (GFAT) plays an integral role in the development of insulin resistance by directing the flow of incoming glucose into the hexosamine biosynthesis pathway. To determine whether the enzymatic activity of GFAT is altered during desensitization of the glucose transport system, we treated isolated rat adipocytes with various combinations of insulin, glucose, and glutamine. Treatment with insulin or glucose alone (or in combination) failed to reduce cytosolic GFAT activity after 4 h, whereas combined treatment with all three components elicited a progressive loss of GFAT activity that was rapid (tH of 2 h) and extensive (70% loss). A pronounced loss of GFAT activity was also seen in cells exposed to glucosamine, an agent known to di- rectly enter the hexosamine pathway (55% loss at 4 h, ED6,, of 360 WM). Moreover, a close correlation was observed between the induction of desensitization and the loss of GFAT activity as a function of glucose, insulin, glutamine, and glucosamine concentrations. When total intracellular hexosamine products were measured, we found that hexosamine formation was unaltered by insulin or glucose (or a combination) but was elevated by >4-fold in the presence of insulin, glucose, and glutamine (tH of 22 min), a condition known to cause both desensitization and loss of GFAT activity. Additional studies indicated that the loss of GFAT activity under desensitizing conditions is not

We reported previously that glutamine:F-6-P amidotransferase (GFAT) plays an integral role in the development of insulin resistance by directing the flow of incoming glucose into the hexosamine biosynthesis pathway. To determine whether the enzymatic activity of GFAT is altered during desensitization of the glucose transport system, we treated isolated rat adipocytes with various combinations of insulin, glucose, and glutamine. Treatment with insulin or glucose alone (or in combination) failed to reduce cytosolic GFAT activity after 4 h, whereas combined treatment with all three components elicited a progressive loss of GFAT activity that was rapid (tH of 2 h) and extensive (70% loss). A pronounced loss of GFAT activity was also seen in cells exposed to glucosamine, an agent known to directly enter the hexosamine pathway (55% loss at 4 h, ED6,, of 360 WM). Moreover, a close correlation was observed between the induction of desensitization and the loss of GFAT activity as a function of glucose, insulin, glutamine, and glucosamine concentrations. When total intracellular hexosamine products were measured, we found that hexosamine formation was unaltered by insulin or glucose (or a combination) but was elevated by >4-fold in the presence of insulin, glucose, and glutamine (tH of 22 min), a condition known to cause both desensitization and loss of GFAT activity. Additional studies indicated that the loss of GFAT activity under desensitizing conditions is not due to allosteric regulation since removal of potential allosteric factors from the cytosol of desensitized cells by G-25 column chromatography failed to restore enzyme activity. Overall, these studies indicate that 1) GFAT is an insulin-regulated enzyme; however, control of enzyme activity is not due to a direct action of insulin, but rather is mediated by insulin-induced enhancement of glucose uptake; 2) the routing of incoming glucose through the hexosamine pathway and the formation of hexosamine products appears to regulate GFAT activity; and 3) the progressive loss of GFAT activity over several hours is probably not due to allosteric regulation.
Desensitization of the insulin-responsive glucose transport system (GTS)' in primary cultured adipocytes requires three components, insulin, glucose, and glutamine (1-5). To address the fundamental question of why glutamine is required for the expression of glucose-induced desensitization, we recently tested the hypothesis that formation of hexosamine products is involved in the induction of insulin resistance (6). Two approaches were used to test this hypothesis. First, we assessed whether glucose-induced desensitization of the GTS could be prevented by glutamine analogs that irreversibly inactivate glutamine-requiring enzymes, such as glutamine:fructose-6-phosphate amidotransferase (GFAT), the first and the rate-limiting enzyme in hexosamine biosynthesis. Both 0-diazoacetyl-L-serine (azaserine) and 6-diazo-5-0~0norleucine inhibited desensitization in 18-h treated cells without affecting maximal insulin responsiveness in control cells. Moreover, a close agreement was seen between azaserine's ability to prevent desensitization of the GTS in intact adipocytes, its ability to inactivate GFAT in intact adipocytes, and its ability to inactivate GFAT activity in a cytosolic adipocyte preparation. From these results, we concluded that a glutamine amidotransferase is involved in the induction of insulin resistance. As a second approach, we determined whether glucosamine, an agent known to preferentially enter the hexosamine pathway at a point distal to enzymatic amidation by GFAT, could induce cellular insulin resistance. These studies revealed that glucosamine can enter adipocytes through the insulin-responsive GTS and effectively desensitize the GTS in a dose-dependent manner (ED,, of 360 FM). In addition, we found that glucosamine was >40 times more potent than glucose in mediating desensitization, did not require glutamine for its desensitizing action, and was able to induce desensitization in the presence of azaserine.
Collectively, these studies indicated that a unique metabolic pathway exists in adipocytes that mediates desensitization of the insulin-responsive GTS and that an early step in this pathway involves the conversion of fructose 6-phosphate to glucosamine 6-phosphate by GFAT. In the current study, we have examined whether GFAT activity in adipocytes is altered under desensitization conditions. The obtained results support this idea and reveal that GFAT activity is regulated in a coordinated manner by insulin, glucose, and glutamine.

EXPERIMENTAL PROCEDURES
Materials-Sources of materials were as follows: porcine monocomponent insulin, Dr. Ronald Chance, Eli Lilly Co., Indianapolis, IN; The abbreviations used are: GTS, glucose transport system; GFAT, glutamine:fructose-6-phosphate amidotransferase; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; HBSS, Hepesbuffered salt solution containing 25 mM Hepes and 1% bovine serum albumin, pH 7.4; MIR, maximum insulin responsiveness; UDP-GlcNAc, UDP-A-acetylglucosamine; F-6-P, fructose 6-phosphate. Preparation of Sterile Isolated Adipocytes-Male Sprague-Dawley rats weighing 160-225 g were killed by cervical dislocation, and the epididymal fat pads were removed under sterile conditions. Isolated adipocytes were then obtained by collagenase digestion (7) as described previously (8,9). Briefly, minced tissue (1-2 g) in 4 ml of Dulbecco's modified Eagle's medium containing 25 mM Hepes, collagenase (1.5 mg/ml), and albumin (40 mg/ml) was shaken in 128 ml sterile polypropylene containers at 37 "C for 50 min. At the end of the digestion period, cells were filtered through nylon mesh (1,000 pm), centrifuged a t 100 rpm for 25 s, and then washed in sterile glucose-free and insulin-free Hepes-buffered balanced salt solution (HBSS) consisting of 20 mM Hepes, 120 mM NaC1, 1.2 mM MgS04, 2 mM CaCls, 2.5 mM KCl, 1.0 mM NaH,PO,, 1 mM sodium pyruvate, 20 units/ml penicillin, 20 mg/ml streptomycin, and 1% bovine serum albumin, pH 7.4. This washing procedure was repeated three times. Since isolated adipocytes are unique from other cell types in that they float on top of the buffer or culture medium, washing of freshly isolated or cultured cells was accomplished by first aspirating the infranatant with a sterile disposable pipet (or a syringe with a 16gauge needle) and then re-adding fresh buffer. Freshly isolated adipocytes were diluted to a final volume equal to 10 ml of buffer/l g of fat (final concentration -5 X IO5 cells/ml) in HBSS.
Measurement of Glucose Transport-Basal and maximally insulinstimulated rates of glucose transport were determined by preincubating adipocytes (0.2 ml in 12 X 75-mm polystyrene tubes) in the absence (basal uptake) or presence of 25 ng/ml insulin for 30 min a t 37 "C (maximal insulin responsiveness). Initial rates of glucose uptake were measured by adding 20 pl of HBSS containing 0.12 pCi of 2deoxy-~-["H]glucose and 2-deoxyglucose (final substrate concentration of 50 p~) .
At the end of 3 min, the reaction was terminated by transferring cells (180 pl) to plastic microcentrifuge tubes containing 60 p1 of silicone oil and centrifuging tubes a t 11,000 X g for 30 s. Since silicone oil has a specific gravity intermediate between buffer and cells, three layers are formed after centrifugation: adipocytes on the top, oil in the middle, and buffer on the bottom (containing free 2-deoxy-~-[:'H]glucose). The cell plug was then removed by cutting the microfuge tube with a razor blade (through the oil layer) and transferring cells to a scintillation vial. After adding a scintillation cocktail, cell-associated radioactivity was determined. T o correct the 2-deoxyglucose uptake values for uptake of hexose by simple diffusion and for nonspecific trapping of radioactivity in the cell pellet, we assessed glucose uptake in the presence of 0.3 mM phloretin (10). In each experiment, glucose uptake was derived from the mean of duplicate or triplicate determinations, sampled in duplicate.
The assay is based on the principle that 2-deoxyglucose is transported and phosphorylated by the same process as D-glucose, but cannot be further metabolized.
Incubation of Isolated Adipocytes and Washing Procedures-In experiments assessing GFAT activity in the cell cytosol or measuring the intracellular concentration of total hexosamine products, freshly isolated adipocytes were added to sterile 50-ml polystyrene tubes (-7 X 10" cells/lO ml for the GFAT assay; -2.1 X lo6 cells/lO ml for assessment of hexosamine levels). After adding various combinations of glucose (20 mM), insulin (25 ng/ml), amino acids (1 X), or glutamine (16 mM), adipocytes were incubated a t 37 "C for various times and then washed. When cytosolic GFAT activity was to be measured, cells were washed three times with ice-cold 50 mM Hepes, pH 7.5, containing 50 mM KC1, 100 mM KH2P04. When total hexosamine products were to be assayed, cells were washed three times with icecold 10 mM Hepes, pH 7.4. For experiments assessing rates of glucose uptake, adipocytes (-1.4 X IO5 cells/2 ml) were added to 17 x 100mm sterile polystyrene tubes and incubated a t 37 "C with various combinations of glucose, insulin, or glutamine. After incubation, control and treated cells were washed three times at 37 "C with glucose-and insulin-free HBSS containing 0.4% bovine serum albumin and then incubated for at least 30 min at 37 "C to allow the GTS to deactivate to near basal levels (95% deactivation). During the final wash, cells were resuspended in HBSS buffer containing 1% bovine serum albumin, pH 7.4, and concentrated to about 2 X lo5 cells/ml. All groups of cells were initially derived from a common pool of freshly isolated cells, were subjected to an equal number of washes, and were concentrated to the same final volume with buffer. Thus, any loss of cells due to washing was identical among control and treated cells. It should also be mentioned that 1 X amino acids is equivalent to the number and concentration of the 15 amino acids found in Dulbecco's modified Eagles medium.
Determination of Total Intracellular Hexosamine Levels-Following treatment, adipocytes were transfered to ice-cold 12 X 75-mm glass tubes and rapidly washed three times with 1 ml of ice-cold 10 mM Hepes, pH 7.4. After the final wash, 500 pl of 6% perchloric acid was added and the mixture was vortexed for 30 s to lyse cells. To neutralize the perchloric acid, 92 pl of ice-cold 6 N potassium hydroxide was added, and the mixture was allowed to sit on ice for 15 min. The tubes were then centrifuged a t 600 X g for 5 min, which resulted in the formation of three distinct layers: congealed fat on top, deproteinated cytosol in the middle, and a pellet containing denatured protein and potassium perchlorate on the bottom. The deproteinated cytosol was removed and analyzed for total hexosamine levels using a modification of the Morgan and Elson method as described by Ghosh et al. (11). Briefly, 40 p1 of sodium tetraborate (800 mM, pH 9.2) and 10 p1 of ice-cold 1.5% aqueous acetic anhydride solution were added to 50 p1 of deproteinated cytosol. The mixture was then heated to 30 "C for 1 min, transferred to a boiling water bath for 3 min to stop the acetylation reaction, and then cooled to 20 "C in a waterbath. Ehrlich's reagent was added to each sample (150 pl, final volume of 250 p l ) , and the mixture was heated for 15 min at 37 "C. After cooling samples to 20 "C, absorbance was recorded a t 585 nm. Blanks were treated identically, except the borate and acetic anhydride solutions were replaced with water. The concentration of hexosamine products in the assay was obtained from standard glucosamine 6-phosphate concentration curves. Under these conditions glucosamine, glucosamine 6-phosphate, and N-acetylglucosamine all yielded identical absorbance values (see Table I). Ehrlich's reagent was prepared by adding 0.4 gm of p-dimethylaminobenzaldehyde to 1 ml of 10 N HCI and then diluting the mixture to 20 ml with glacial acetic acid.
Measurement of Glutamine:Fructose-6-phosphateAmidotransferase Actiuity in Adipocytes-Cells (7 X 105 cells) were suspended in lysis buffer containing 50 mM Hepes, 100 mM KH2P04, and 50 mM KCI, pH 7.5, and mechanical breakage of adipocytes was performed by vortexing 1 ml of cells in 1 2 X 75-mm glass tubes for 2 min at 4 "C. The tubes were then placed on ice for 10 min, after which time 800 p1 of the infranatent (the aqueous portion of the mixture below the congealed fat layer) was transferred by syringe to 1.5-ml microcentrifuge tubes and centrifuged for 1 min a t 13,000 X g. Tubes were again cooled on ice, 700 p1 of infranatant was transferred to another 1.5 ml of microcentrifuge tube, and the microcentrifugation step was repeated. In the last step, 600 p1 of infranatent was withdrawn and transferred to ice-cold 12 X 75-mm tubes. Measurement of GFAT was performed using this crude cytosol preparation.
Glutamine:fructose-6-phosphate amidotransferase (EC 2.6.1.16) was measured using a spectrophotometric assay, essentially as described by Shijo et al. (12). In a typical assay, a cuvette containing 10 mM fructose 6-phosphate, 6 mM glutamine, 0.3 mM 3-acetylpyridine adenine dinucleotide, 50 mM KC1, 100 mM KH,PO,, and 6 units of glutamate dehydrogenase was equilibrated for 30 min at 30 "C (1 ml total assay volume, p H 7.5). After initiating the reaction by adding 200 p1 of cytosol, the change in absorbance caused by reduction of 3acetylpyridine adenine dinucleotide was monitored spectrophotometrically a t 365 nm. Under these conditions, the activity remained linear for at least 30 min. The enzyme activity was calculated by assuming c:lBfi (3-acetylpyridine adenine dinucleotide) = 0.91 liter. mm".mmol" (13), and a unit of activity was defined as 1 nmol of product formed per min. GFAT activity in the crude cytosol preparation remained constant a t 4 "C for at least 6 h, and all enzyme assays were performed under optimal substrate and linear kinetic conditions. Values were normalized to cytosolic protein concentrations which were measured using a Pierce Chemical Co. bicinchoninic acid protein assay kit.
Definition of Terms-Based on the finding that loss of GFAT activity requires insulin, glucose, and glutamine, we variously refer to GFAT as an either an insulin-regulated enzyme or a glucoseregulated enzyme. The choice of terms reflects a matter of perspective and is used to conceptually facilitate discussion by focusing on a single variable. For example, when discussing regulation of GFAT with regard to other insulin-regulated enzymes, we refer to GFAT as an insulin-regulated enzyme. Implicit in this term, however, is the fact that glucose and glutamine are also present at maximally effective concentrations. Alternatively, when discussing the idea that the up-take and routing of glucose through the hexosamine biosynthesis pathway plays an integral role in regulating GFAT, we use the term glucose-regulated enzyme. Last, we sometimes refer to GFAT as a hexosamine-regulated enzyme to reflect the idea that formation of hexosamine products is linked to the regulation of GFAT activity.

Individual and Combined Effects of Glucose, Insulin, and
Glutamine on GFATActiuity-Shown in Fig. 1B is an experiment in which isolated adipocytes were suspended in HBSS containing 20 mM glucose and 1 X amino acids and treated with 25 ng/ml insulin to induce desensitization of the GTS. At the indicated times, cells were thoroughly washed and lysed to obtain a crude cytosolic preparation. When GFAT activity was then measured, we observed a rapid loss of GFAT activity ( t M of 2 h) that plateaued after 4 h to a new steadystate level that was 60-70% below control values. The insulininduced loss of GFAT activity was specific in that GFAT activity in control cells remained constant during the 5-h incubation. As can be seen in Fig. lA, insulin-induced desensitization was temporally similar to the loss of GFAT activity, although slightly slower (tlh of 3 h).
Since we demonstrated previously that desensitization requires simultaneous treatment with insulin, glucose, and glutamine (6), we next assessed whether loss of GFAT activity similarly requires all three components. As can be clearly seen in Fig. 2, when cells were incubated for 4 h in HBSS containing various combinations of glucose (20 mM), insulin (25 ng/ ml), and glutamine (16 mM), we found that neither glucose alone nor insulin alone (or a combination of the two) affected GFAT activity, whereas combined treatment with all three components elicited a significant reduction in enzyme activity. The dose-response curves shown in Fig. 3 1. Ability of insulin to induce loss of glutamine:fructose-6-P amidotransferase activity and mediate desensitization of the glucose transport system. A , isolated adipocytes were suspended in HBSS containing 20 mM glucose and 1 X amino acids and incubated a t 37 "C in the absence (controls) or presence of 25 ng/ml insulin. (1 x amino acids is equivalent to number and concentration of amino acids found in Dulbecco's modified Eagles medium and contains 584 mg of glutamine/liter). At the indicated times, cells were washed and glutamine:fructose-6-P amidotransferase activity was measured in a crude cytosol preparation as described in the methods section. B, after treating adipocytes as described above, cells were washed, preincubated a t 37 "C for 30 min with 25 ng/ml insulin, and assayed to determine maximally insulin-stimu-   concentration of 20 mM), only D-mannose effectively substituted for D-glucose in lowering GFAT activity. These results agree well with our previous studies (5,14) showing that desensitization of the GTS occurs in the presence of D-glucose and D-mannose, but not D-galactose, D-fructose, or 3-0-methylglucose.
Effects of Glucosamine on GFAT Actiuity-On the basis of the similarities between the loss of GFAT and the induction of desensitization (Figs. 1-3), we postulated that products of the hexosamine biosynthesis pathway coordinately regulate both desensitization and the loss of GFAT activity. We tested this idea by examining the effects of glucosamine and insulin treatment on GFAT activity. The rationale behind this pharmacological approach is based on the fact that glucosamine is known to directly enter the hexosamine biosynthesis pathway (15-18) where it rapidly mediates induction of insulin resistance (6). As depicted in Fig. 5A, when cells were treated for 4 Under similar conditions, glucosamine treatment led to desensitization of the GTS with an EDso of 360 PM (Fig. 5A,   inset).
The data presented in Fig. 5 B highlight two important aspects of glucosamine-induced loss of GFAT activity. First, insulin is required for glucosamine to lower GFAT activity, which is not unexpected since insulin enhances glucosamine uptake into adipocytes by about 10-fold (6). Second, glutamine is not required for glucosamine-induced loss of GFAT activity, which is readily explained by the fact that glucosamine directly enters the hexosamine biosynthesis pathway a t a point distal to GFAT action. Thus, glutamine is not required as an amide donor. Overall, these studies add credence to the idea that the flow of glucose through the hexosamine biosynthesis pathway coordinately regulates both the insulin-responsive GTS and the enzymatic activity of GFAT.
Individual and Combined Effects of Glucose, Insulin, and Glutamine on the Formation of Hexosamine Products-Shown in Fig. 6 A is an experiment in which adipocytes were preincubated with 25 ng/ml insulin and 16 mM glutamine (37 "C for 30 min) and then exposed to 20 mM glucose. At the indicated times, cells were washed, lysed, and assayed to determine the intracellular levels of total hexosamine products as described under "Experimental Procedures." As can be seen, hexosamine products rapidly increased after the addition of glucose (tzfi2 of 22 min) and plateaued between 1 and 2 h at a level >4-fold higher than control cells. After 2 h, a slow decline in total hexosamine products was observed which could be explained by the progressive loss of GFAT activity under these desensitizing conditions (Fig. 1A).
When the individual and combined effects of glucose, insulin, and glutamine on hexosamine levels were examined after 1 h (Fig. 6 B ) , we found that enhanced hexosamine levels were observable only when all three components were present. Moreover, we found that the intracellular hexosamine levels were significantly elevated in cells treated with 1 mM glucosamine and insulin. Both of these findings add credence to the idea that formation of hexosamine products is intimately involved in both the loss of GFAT activity and desensitization of the GTS (Fig. 1). It is also noteworthy that an extracellular source of glutamine was required for hexosamine production. Since glutamine is presumably serving as an amide donor, this finding indicates that endogenous protein catabolism in adipocytes is unable to provide sufficient glutamine for the normal functioning of GFAT. Table I summarizes the results of experiments in which the specificity of the hexosamine assay was determined by using a known amount (100 p~) of commercially available hexosamine products. The major finding of this study is that the employed hexosamine assay measures several hexosamine products rather than a single product. However, several other interesting findings are also apparent. For example, in general we found that early hexosamine products are the most reactive (glucosamine-6-P and N-acetylglucosamine-6-P), whereas later products are the least reactive. Additionally, hexosamines phosphorylated in the 1 position rather than the 6 FIG. 6. Formation of total hexosamine products under desensitizing conditions. A , cells were preincubated for 30 min at 37 "C with insulin (25 ng/ml) and glutamine (16 mM) and then treated with 20 mM glucose. At the indicated times, adipocytes were lysed and the concentration of intracellular hexosamine products was measured as described in the methods section. Each point represents the mean t S.E. of three experiments. B, adipocytes were treated for 1 h a t 37 "C with the indicated components, lysed, and assayed to determine total hexosamine products. Each bar represents the mean f S.E. The n number is shown in white at the bottom of each bar. The total amount of hexosamine products in the presence of glucose alone was compared to that found in other treatment groups using the Student's t test.. ***, p < 0.001.

TABLE I
Specificity of hexosamine assay Hexosamine measurements were conducted as described under "Experimental Procedures" using 100 pM of commercially obtained hexosamine products. Data represent the mean f S.E. of two experiments. Effects of UDP-N-Acetylglucosamine on GFAT Activity-GFAT derived from liver and other mammalian tissues has been shown to be allosterically regulated by UDP- GlcNAc  (19-22). To confirm that the adipocyte form of GFAT is similarly regulated by UDP-GlcNAc, we performed the experiment shown in Fig. 7. Using a crude adipocyte cytosolic preparation as our source of enzyme, we measured GFAT activity in the presence of 10 mM F-6-P, 6 mM glutamine, and varying concentrations of UDP-GlcNAc (Fig. 7A). As can be clearly seen, UDP-GlcNAc lowered adipocyte GFAT activity in a dose-dependent manner (ED50 of 52 pM).

Compound Absorbance
The finding that UDP-GlcNAc negatively regulates GFAT activity under in vitro conditions naturally raises the question of whether the loss of GFAT activity we observed under desensitization conditions could be due to carry-over of intracellular UDP-GlcNAc into the in vitro enzyme assay. As a first step in addressing this question, we used [14C]3-0-methylglucose and phloretin to measure the intracellular water space in adipocytes (23) and then calculated the final concentration of cytosol in the GFAT assay. These studies revealed that the cytosol was diluted more than 500-fold in the crude cytosol preparation and more than 2,500-fold in the GFAT assay (data not shown). Thus, it appears unlikely that UDP-GlcNAc or any other potential inhibitory factors are involved in the glucose-induced loss of GFAT activity since they would most likely have been diluted beyond the point where inhibition of GFAT activitity could be observed. Nevertheless, to conclusively demonstrate that allosteric regulation is not involved in glucose-induced loss of GFAT activity, we measured GFAT activity in the cytosol from control and 4-h desensitized adipocytes before and after chromatography on a Sephedex G-25 column. As can be seen in Fig. 7B, GFAT activity was markedly reduced in the cytosol derived from desensitized cells as well as in control cytosol to which we added UDP-GlcNAc (500 p~ final concentration). After G-25 chromatography of the cytosol, GFAT activity was restored in the group containing UDP-GlcNAc, but not in the the desensitized cell POUP.
Studies Determining K,,, and Vmx Values for GFAT in Adipocytes-After preparing a crude cytosolic enzyme preparation from control and 4-h desensitized adipocytes, we assessed GFAT activity in the presence of 10 mM F-6-P as a function of increasing glutamine concentration (from 0.1 to 6 mM). When we calculated the average glutamine K,,, values for GFAT from double reciprocal plots of the reaction rate against glutamine Concentration, we found no significant differences between the control and desensitized groups (Table   11). Moreover, the glutamine K,,, values for the adipocyte form of GFAT (about 400 p~) was similar to the reported GFAT values derived from liver and other tissues (0.6-1.6 mM).
The K , for F-6-P was similarly obtained by varying the concentration of F-6-P (0.1 to 10 mM) while holding the concentration of glutamine constant at 6 mM. Again, no significant differences were detected between the control and desensitized groups (Table 11), and the observed values agreed well with the reported GFAT K, (F-6-P) values for other mammalian tissues, which ranges from 200 UM to 500 UM (20, 24-26). In contrast to the invariant K, values in the desensitized group, we did find a significant reduction ( p >0.001) in the VmaX.

DISCUSSION
In our earlier studies, we demonstrated that desensitization of the insulin-responsive GTS requires insulin, glucose, and glutamine (1,2,4,5). More recently, we found that that GFAT plays an integral role in the development of insulin resistance by directing the flow of incoming glucose into the hexosamine biosynthesis pathway (6). The current study extends this line of investigation by revealing that GFAT can be hormonally regulated by insulin. For example, when isolated rat adipocytes were suspended in HBSS containing glucose and amino acids and treated for various times with 25 ng/ml insulin (Fig. l), we observed a progressive loss of GFAT activity that was both rapid (tlh of 2 h) and extensive (70% decrease). Moreover, this effect was specific in that GFAT activity remained constant for up to 5 h in the absence of insulin. Importantly, the ability of insulin to diminish GFAT activity occurred within the insulin dose range known to stimulate glucose uptake and enhance the net rate of protein synthesis (3).
Although the hexosamine biosynthesis pathway has long been recognized as a glucose-utilizing pathway, there have been no reports in the literature identifying insulin-responsive enzymes within this pathway. Therefore, our finding that GFAT is an insulin-regulated enzyme is significant in that it extends the regulatory actions of insulin to an important anabolic pathway intimately involved in the de nouo formation of glycoproteins, specialized glycolipids, and proteoglycans. In retrospect, this finding is not unusual, since it is well established that insulin is an anabolic hormone that plays a central role in the metabolism of carbohydrates. For example, insulin controls the flux of glucose through the glycogen biosynthesis pathway by regulating glycogen synthase activity (27-30), it controls the flux through the pentose phosphate shunt by influencing glucose-6-phosphate dehydrogenase (31)(32)(33)(34), and it modulates the glycolytic pathway by controlling several enzymes including glyceraldehyde-3-phosphate dehydrogenase (35-38), pyruvate kinase (39-45), and phosphoenylpyruvate carboxykinase (46,46). The finding that insulin regulates the first and rate-limiting enzyme of the hexosamine biosynthesis pathway (GFAT) is a natural extension of its primary role as an anabolic hormone.
The mechanism underlying insulin regulation of GFAT activity is rather unique in that the insulin-induced loss of enzyme activity is not due to a direct action of insulin but rather is mediated by enhanced glucose uptake and the subsequent flux of glucose into the hexosamine biosynthesis pathway. Thus, a 4-h exposure to either insulin alone or glucose alone failed to lower GFAT activity, whereas combined treatment with insulin, glucose, and glutamine elicited a marked reduction in GFAT activity. Treatment of adipocytes with these three components also desensitized the GTS as previously reported (6), and a close correlation was seen between the induction of desensitization and the loss of GFAT activity as a function of glucose, insulin, and glutamine concentrations (similar EDs0 values). Moreover, glucosamine lowered GFAT activity in a dose-dependent manner after a 4-h treatment period (55% loss), and the concentration of glucosamine required for loss of enzyme activity (EDso of 210 p M ) was similar to that causing desensitization of the GTS (EDs0 of 360 pM). The ability of glucosamine to mediate loss of GFAT activity is noteworthy for two reasons, glucosamine has been shown previously to preferentially enter the hexosamine pathway at a point distal to amidation of GFAT (19-22), and glucosamine can readily be transported into adipocytes through the insulin-responsive GTS where it can effectively desensitize the GTS with a potency that is 40-times greater than glucose (6). Overall, these combined studies indicate that both the induction of desensitization and the loss of GFAT activity are tightly coupled to the formation of hexosamine products.
In 1964, Kornfeld et al. (19) first reported that GFAT derived from liver can be allosterically regulated by UDP-Nacetylglucosamine, the immediate precursor for glycoprotein biosynthesis. Subsequent investigators rapidly confirmed this basic observation (20,22,26) and studies by Bates et ai. (26) extended this concept by showing that UDP-N-acetylglucosamine regulates GFAT activity in uiuo (in intact cells). In agreement with these earlier studies, we have found that UDP-N-acetylglucosamine decreases GFAT activity in a cytosolic preparation derived from isolated adipocytes, and the obtained EDso value of 52 ~L M closely agrees with reported Ki values for other tissues (20-70 FM).
Although a t first glance it would appear that the observed loss of GFAT activity in adipocytes (under desensitizing conditions) is due to allosteric regulation, we believe that this is not the case for several reasons. First, we found that loss of GFAT activity occurs over several hours (tH of 2 h) rather than minutes as would be expected for allosteric regulation. Second, using ['4C]-3-O-methylglucose and phloretin to measure the intracellular water space in adipocytes, we calculated that the cytosol in the GFAT assay was diluted greater than 2,500-fold. Thus, it appears unlikely that UDP-N-acetylglucosamine or any other potential allosteric effectors are involved in the hexosamine-induced loss of GFAT activity since they would most likely have been diluted beyond the point where allosteric regulation of GFAT could be observed. Last, the removal of low M, allosteric regulators such as UDP-Nacetylglucosamine from the cytosol of desensitized cells by G-25 column chromotography did not restore diminished GFAT activity to control values. Although it appears that allosteric regulation does not underlie the loss of GFAT, it is important to emphasize that we are not ruling out the idea that UDP-N-acetylglucosamine is a regulator of GFAT activity in intact adipocytes. In fact, our current view is that several different mechanisms may operate to regulate GFAT activity, including a short-term system involving allosteric regulation and possibly a long-term system involving regulation of gene transcription or mRNA stability.
The functional significance of the coordinated regulation of GFAT activity by insulin, glucose and glutamine remains to elucidated; however, it is possible that the loss of GFAT occurring under desensitizing conditions serves to limit the extent to which glucose can induce desensitization of the GTS. Such a negative feedback system would ensure that cells retain at least some degree of responsiveness to circulating levels of insulin. Along these lines, we have found that maximal insulin responsiveness of the GTS cannot be completely desensitized even after prolonged treatment times (24 h) with high concentrations of glucose, insulin, and glutamine. In other words, loss of insulin responsiveness always plateaus at about 20-30% of control values (1,4). Clearly, additional studies will be required to clarify the mechanisms underlying glucose-induced loss of GFAT and the physiological significance of this regulatory process.