Implication of eIF2B Rather Than eIF4E in the Regulation of Global Protein Synthesis by Amino Acids in L6 Myoblasts*

The present study was designed to investigate the mechanism through which leucine and histidine regulate translation initiation in L6 myoblasts. The results show that both amino acids stimulate initiation and coordinately regulate the activity of eukaryotic initiation factor eIF2B. The changes in eIF2B activity could be explained in part by modulation of the phosphorylation state of the α-subunit of eIF2. The activity changes might also be a result of modulation of the phosphorylation state of the eIF2B ε-subunit, because deprivation of either amino acid caused a decrease in eIF2Bε kinase activity. Leucine, but not histidine, additionally caused a redistribution of eIF4E from the inactive eIF4E·4E-BP1 complex to the active eIF4E·eIF4G complex. The redistribution was a result of increased phosphorylation of 4E-BP1. The changes in 4E-BP1 phosphorylation and eIF4E redistribution associated with leucine deprivation were not observed in the presence of insulin. However, the leucine- and histidine-induced alterations in global protein synthesis and eIF2B activity were maintained in the presence of the hormone. Overall, the results suggest that both leucine and histidine regulate global protein synthesis through modulation of eIF2B activity. Furthermore, under the conditions employed herein, alterations in eIF4E availability are not rate-controlling for global protein synthesis but might be necessary for regulation of translation of specific mRNAs.

The present study was designed to investigate the mechanism through which leucine and histidine regulate translation initiation in L6 myoblasts. The results show that both amino acids stimulate initiation and coordinately regulate the activity of eukaryotic initiation factor eIF2B. The changes in eIF2B activity could be explained in part by modulation of the phosphorylation state of the ␣-subunit of eIF2. The activity changes might also be a result of modulation of the phosphorylation state of the eIF2B ⑀-subunit, because deprivation of either amino acid caused a decrease in eIF2B⑀ kinase activity. Leucine, but not histidine, additionally caused a redistribution of eIF4E from the inactive eIF4E⅐4E-BP1 complex to the active eIF4E⅐eIF4G complex. The redistribution was a result of increased phosphorylation of 4E-BP1. The changes in 4E-BP1 phosphorylation and eIF4E redistribution associated with leucine deprivation were not observed in the presence of insulin. However, the leucine-and histidine-induced alterations in global protein synthesis and eIF2B activity were maintained in the presence of the hormone. Overall, the results suggest that both leucine and histidine regulate global protein synthesis through modulation of eIF2B activity. Furthermore, under the conditions employed herein, alterations in eIF4E availability are not ratecontrolling for global protein synthesis but might be necessary for regulation of translation of specific mRNAs.
Amino acids play important and multiple roles in regulating protein synthesis in skeletal muscle (1). An obvious role is to act as precursors for protein synthesis. A less obvious but equally important role involves the regulation of translation initiation. The initiation of mRNA translation is a complicated process involving over a dozen proteins, referred to as eukaryotic initiation factors (eIFs) 1 (reviewed in Refs. 2 and 3). Of all the steps in the initiation pathway, only two have been identified that are subject to regulation in vivo; the binding of tRNA i Met to the 40 S ribosomal subunit and the binding of mRNA to the 43 S preinitiation complex. In the first step in initiation, tRNA i Met binds to the 40 S ribosomal subunit as a ternary complex with eIF2 and GTP. Subsequently, the GTP bound to eIF2 is hydrolyzed to GDP, and eIF2 is released from the ribosomal subunit as a complex with GDP. Formation of the ternary complex is regulated by modulation of the activity of a second initiation factor, eIF2B, which mediates guanine nucleotide exchange on eIF2. It is regulated by phosphorylation of the ␣-subunit of eIF2, where phosphorylation converts eIF2 from a substrate into a competitive inhibitor of eIF2B.
The binding of mRNA to the 43 S preinitiation complex involves a group of proteins collectively referred to as eIF4 (reviewed in Refs. 2 and 4). The protein that binds to the m 7 GTP cap present at the 5Ј-end of most eukaryotic mRNAs is termed eIF4E. The eIF4E⅐mRNA complex binds to the 40 S ribosomal subunit through the association of eIF4E with eIF4G. An important mechanism for regulating the binding of mRNA to the 40 S ribosomal subunit occurs through sequestration of eIF4E into an inactive complex with the eIF4E binding protein, 4E-BP1 (5,6). The eIF4E⅐4E-BP1 complex can still bind to the m 7 GTP cap structure but not to eIF4G (7,8). Thus, binding of eIF4E to 4E-BP1 prevents the binding of mRNA to the ribosome. The binding of 4E-BP1 to eIF4E is regulated by phosphorylation of 4E-BP1 with increased phosphorylation of the protein causing a decrease in the affinity of 4E-BP1 for eIF4E.
In skeletal muscle in vivo as well as in perfused muscle preparations, protein synthesis is stimulated by amino acids through a mechanism involving an increase in the rate of translation initiation (reviewed in Ref. 1). The stimulation of initiation caused by a complete mixture of amino acids can be reproduced by provision of only the branched chain amino acids (9,10). Furthermore, of the branched chain amino acids, leucine has been shown to stimulate initiation in the absence of other amino acids (11), suggesting that it plays a key role in the regulation of protein synthesis. However, although leucine has been known to stimulate protein synthesis in muscle for years, the mechanism through which it acts on initiation is still unknown.
Leucine also stimulates initiation in tissues other than skeletal muscle (1). However, in tissues such as liver, leucine is not the only amino acid that can regulate initiation. In particular, in rat liver, histidine stimulates initiation through a mechanism involving changes in the phosphorylation state of the ␣-subunit of eIF2 (eIF2␣) (12). The decreased phosphorylation of eIF2 caused by histidine is associated with a stimulation of eIF2B activity. Whether leucine has a similar effect on eIF2 phosphorylation in liver or muscle is unknown. In the present study, the mechanism through which leucine and histidine regulate translation initiation was examined in L6 myoblasts.

EXPERIMENTAL PROCEDURES
Materials-ECL detection reagents and horseradish peroxidase-conjugated sheep anti-mouse Ig and donkey anti-rabbit Ig were purchased from Amersham Pharmacia Biotech. Polyvinylidene difluoride membrane was obtained from Bio-Rad. 35 S-Easytag Express protein labeling mix was from NEN Life Science Products. The antibody against p70 S6k was purchased from Santa Cruz Biotechnology, Inc. DMEM lacking either leucine or histidine was purchased from Life Technologies, Inc. Glycogen synthase kinase-3 (GSK-3) substrate and control peptides were purchased from Upstate Biotechnology, Inc.
L6 Myoblast Culture-L6 myoblast cells were grown in culture in 100-mm dishes in DMEM supplemented with 10% fetal bovine serum (Hyclone Labs, Inc), 100 units/ml benzylpenicillin, and 100 g/ml streptomycin sulfate. Cells were grown to approximately 70% confluence and washed twice with phosphate-buffered saline; dishes were then randomly divided into three groups. Dishes in the first group received serum-free DMEM, the second group received serum-free DMEM lacking leucine, and the third group received serum-free DMEM lacking histidine. The cells were returned to the incubator for 60 min, at which time insulin was added to some of the dishes in each group to a final concentration of 20 nM. This concentration of the hormone was found in an earlier study to be maximally effective in activating p70 S6k (13). In addition, leucine was added to some of the dishes in group 2, and histidine was added to some of the dishes in group 3. In each case, the final concentration of leucine or histidine was the same as in complete DMEM. The cells were returned to the incubator, and 30 min later all dishes received 10 l of 35 (14).
Measurement of eIF2B Activity-Cells (approximately 2 ϫ 10 6 ) were washed with ice-cold phosphate-buffered saline and then lysed in 400 l of ice-cold buffer B (45 mM HEPES, pH 7.4, 0.375 mM magnesium acetate, 0.075 mM EDTA, 95 mM potassium acetate, 2.5 mg/ml digitonin, and 10% glycerol). The homogenates were centrifuged at 10,000 ϫ g for 10 min at 4°C, and the supernatants were assayed for the exchange of [ 3 H]GDP bound to eIF2 for unlabeled GDP as described previously (15).
Measurement of eIF2B⑀ Kinase Activity in L6 Myoblasts-L6 myoblasts were washed twice with 3.0 ml of ice-cold phosphate-buffered saline and then lysed using a Dounce homogenizer in buffer consisting of 25 mM Tris⅐HCl, pH 7.0, 0.3 mM ATP (42 Ci/mol), 10 mM magnesium chloride, 5% glycerol, 4.4 mM ␤-mercaptoethanol, 0.5 mM benzamidine, and 0.03 mM phenylmethylsulfonyl fluoride. The homogenate was centrifuged at 10,000 ϫ g for 10 min at 4°C. Aliquots of the supernatant (20 l) were added to tubes containing 30 l of a solution consisting of 1.5 g of recombinant eIF2B⑀ (expressed in and purified from Sf9 insect cells using the baculovirus expression system as described previously (16)), and (final concentrations) 25 mM Tris⅐HCl, pH 7.0, 0.3 mM ATP (43 Ci/mol), 10 mM magnesium chloride, 5% glycerol, 4.4 mM ␤-mercaptoethanol, 0.5 mM benzamidine, and 0.03 mM phenylmethylsulfonyl fluoride. The reaction mixture was incubated at 30°C, and at various times 10 l of the reaction mixture were removed and placed in a separate tube containing 10 l of SDS sample buffer at 90°C. The samples were incubated at 90°C for 4 min and then were resolved by electrophoresis on a SDS-polyacrylamide gel as described previously (17). The stained gel was dried and exposed to film. The results were analyzed by scanning densitometry.
Measurement of GSK-3 Activity in L6 Myoblast Extracts-L6 myoblast homogenates were prepared and centrifuged as described in the previous paragraph, except that the homogenization buffer consisted of 8 mM MOPS, pH 7.4, 0.2 mM EDTA, 10 mM magnesium acetate, 0.1 mM phenylmethylsulfonyl fluoride, and 0.5 mM benzamidine. Aliquots (12.5 l) of the supernatant were assayed for GSK-3 activity at 30°C in a 50-l reaction consisting of 4 mM MOPS, pH 7.4, 0.1 mM EDTA, 20 mM magnesium chloride, 250 M [ 32 P]ATP (1.6 Ci/mol), and 62.5 M of either control or substrate peptide. The control peptide is identical to the substrate peptide except that the serine at the GSK-3 phosphorylation site was changed to an alanine. At various times, 10 l of the reaction mixture were removed and placed in a tube on ice containing 10 l of stop buffer (2% bovine serum albumin, 20 mM ATP, and 1 M HCl). Ice-cold 25% trichloroacetic acid (5 l) was added to each tube followed by incubation on ice for 5 min. The tubes were then centrifuged at 8500 ϫ g for 3 min at 4°C, and 20 l of supernatant were spotted onto 2.1-cm P81 filter disks (Whatman). The filters were incubated in ice-cold 0.75% phosphoric acid (5 ml/filter) for 5 min. The filters were then washed consecutively with ice-cold 0.75% phosphoric acid (twice), water, and 95% ethanol (twice). The filters were dried and radiation bound to the filters was quantitated by liquid scintillation spectrometry. GSK-3 activity was calculated as the difference in the amount of 32 P i incorporated into the substrate compared with the control peptide. Protein Immunoblot Analysis-Blots were developed using an Amersham ECL Western blotting kit as described previously (17). Films were scanned using a Microtek ScanMaker III scanner equipped with a transparent media adapter connected to a Macintosh PowerMac 7100 computer. Images were obtained using the ScanWizard Plugin (Microtek) for Adobe Photoshop and quantitated using NIH Image software.
Determination of eIF2␣ Phosphorylation State-L6 myoblasts were maintained in culture as described above with the exception that the cells were grown on 60 mm, instead of 100 mm, dishes. In addition, cells were harvested by scraping in SDS sample buffer at 90°C as described previously (18). The phosphorylation state of eIF2␣ then was determined by two different methods. In the first method, the relative amount of eIF2␣ in the phosphorylated form was quantitated by protein immunoblot analysis using an affinity-purified antibody that specifically recognizes eIF2␣ phosphorylated at Ser 51 (eIF2(␣P)) (Ref. 19; kindly provided by Drs. Gary S. Krause and Donald J. DeGracia, Wayne State University School of Medicine). For this analysis, samples were resolved by SDS-polyacrylamide gel electrophoresis on a 12.5% polyacrylamide gel, and the proteins in the gel were electrophoretically transferred to a polyvinylidene difluoride membrane as described previously (17). The membranes were incubated with the anti-eIF2(␣P) antibody, and the blots were developed using an ECL Western blot kit as described above. The horseradish peroxidase coupled to the antirabbit secondary antibody was then inactivated by incubating the blot in 15% H 2 O 2 for 30 min at room temperature. The total amount of eIF2␣ in the samples was determined by reprobing the blot with a monoclonal antibody (kindly provided by Dr. Richard Panniers, National Institutes of Health) that equally recognizes the phosphorylated and unphosphorylated forms of eIF2␣ (20) followed by an anti-mouse secondary antibody. Values obtained using the anti-eIF2(␣P) antibody were normalized for the total amount of eIF2␣ present in the sample.
In the second method, the proportion of eIF2␣ present in the phosphorylated form was determined by protein immunoblot analysis following separation of the phosphorylated and unphosphorylated forms of the protein using slab gel isoelectric focusing electrophoresis. Aliquots of the samples in SDS sample buffer were heated at 100°C for 3 min, cooled to room temperature, and then mixed with 0.8 volumes of isoelectric focusing gel buffer (0.1 g dithiothreitol, 0.4 g of CHAPS, 5.4 g of urea, and 1 ml of ampholytes (pH 3.5-9.5 from Amersham Pharmacia Biotech) in 6 ml of water). Samples were resolved by isoelectric focusing followed by protein immunoblot analysis using the monoclonal anti-eIF2␣ antibody as described previously (21).
Quantitation of 4E-BP1⅐eIF4E and eIF4G⅐eIF4E Complexes-The association of eIF4E with 4E-BP1 or eIF4G was quantitated by a modification of the method described previously (22). Briefly, eIF4E and the 4E-BP1⅐eIF4E and eIF4G⅐eIF4E complexes were immunoprecipitated from aliquots of cell homogenate using the anti-eIF4E monoclonal antibody. The eIF4E antibody was raised against recombinant human eIF4E as described previously (23). The efficiency of eIF4E immunoprecipitation was greater than 90% under all conditions. The antibody⅐ antigen complex was collected by incubation for 1 h with goat antimouse Biomag IgG beads (PerSeptive Diagnostics). Before use, the beads were washed in 1% nonfat dry milk in buffer C (20 mM Tris⅐HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 0.1% ␤-mercaptoethanol, 0.5% Triton X-100). The beads were captured using a magnetic stand and washed twice with buffer C and once with buffer D (50 mM Tris⅐HCl, pH 7.4, 500 mM NaCl, 5 mM EDTA, 0.04% ␤-mercaptoethanol, 1% Triton X-100, 0.5% sodium deoxycholate, and 0.1% sodium dodecylsulfate). Protein bound to the beads was eluted by resuspending the beads in SDS sample buffer and boiling the sample for 5 min. The beads were collected by centrifugation, and the supernatants were subjected to electrophoresis either on a 7.5% polyacrylamide gel for quantitation of eIF4G or on a 15% polyacrylamide gel for quantitation of 4E-BP1 and eIF4E. Proteins were then electrophoretically transferred to a polyvinylidene difluoride membrane as described previously (17). The membranes were incubated with a mouse anti-human eIF4E antibody, a rabbit anti-rat 4E-BP1 antibody, or a rabbit anti-human eIF4G anti-body for 1 h at room temperature. The antibodies against 4E-BP1 and eIF4G were raised against the recombinant rat and human proteins, respectively, by the method described previously (24). The blots then were developed using an ECL Western blotting kit as described above.
Examination of 4E-BP1 Phosphorylation in Extracts of L6 Myoblasts-Aliquots of cell homogenates were immunoprecipitated using a monoclonal anti-4E-BP1 antibody using the method described in the previous section for immunoprecipitation of eIF4E. The 4E-BP1 anti-body was raised against the rat recombinant protein by the method described previously (17). The immunoprecipitates were solubilized with SDS sample buffer and then subjected to protein immunoblot analysis using a rabbit anti-rat 4E-BP1 antibody.
p70 S6k Phosphorylation-An aliquot of L6 cell homogenate was combined with an equal volume of SDS sample buffer, and the diluted samples were subjected to electrophoresis on a 7.5% polyacrylamide gel (25). The samples were then analyzed by protein immunoblot analysis using rabbit anti-rat p70 S6k polyclonal antibodies as described above.

RESULTS
The effect of leucine or histidine deprivation on the rate of global protein synthesis in L6 myoblasts was examined by incubating cells in serum-free medium lacking either leucine or histidine for 60 min followed by readdition of the deprived amino acid for 30 min. The results were compared with cells maintained in medium with a complete mixture of amino acids. As shown in Fig. 1A, deprivation of either amino acid caused a reduction in protein synthesis to approximately 50% that of the rate observed in cells maintained in complete serum-free medium. Cells maintained in serum-free DMEM containing all amino acids will be referred to hereafter as controls. Readdition of the deprived amino acid rapidly returned protein synthesis to control values.
In a previous study using perfused rat liver, the inhibition of protein synthesis caused by histidine deprivation was attributed to a reduction in the guanine nucleotide exchange activity of eIF2B (12). In the present study, deprivation of either leucine or histidine also caused a significant decline in the guanine nucleotide exchange activity of eIF2B in L6 myoblasts (Fig. 1B). Readdition of the deprived amino acid rapidly restored eIF2B activity to the value observed in control cells.
In both mammalian cells (12) and yeast (26), the inhibition of eIF2B activity caused by deprivation of histidine is associated with an increase in phosphorylation of eIF2␣. To determine whether leucine or histidine deprivation caused a similar effect in L6 myoblasts, cell homogenates were subjected to protein immunoblot analysis using an antibody that recognizes eIF2␣ only when it is phosphorylated at Ser 51 (19). As shown in Fig.  1C, the amount of phosphorylated eIF2␣ increased approximately 2.5-fold in response to deprivation of either leucine or histidine compared with the amount in control cells. Readdition of the deprived amino acid completely reversed the increased phosphorylation of eIF2(␣P). However, as described below, the amount of eIF2␣ in the phosphorylated form represents a relatively low proportion of the total protein.
The mechanism through which eIF2␣ phosphorylation causes an inhibition of eIF2B activity involves an increased affinity of eIF2B for phosphorylated compared with unphosphorylated eIF2 (reviewed in Ref. 2). The increased affinity of FIG. 1. Effects of leucine or histidine deprivation and resupplementation on protein synthesis, eIF2B activity, and phosphorylation of eIF2␣ in L6 myoblasts. A, L6 myoblasts were grown in culture as described under "Experimental Procedures" and were then randomly divided into three groups. The results are expressed as a percent of the maximum control value within a given experiment and represent the mean ϮS.E. of 16 -18 experiments. Within each experiment, 2-3 cultures/condition were individually analyzed. In each panel, hatched bars represent cells deprived of either leucine or histidine as indicated in the figure; black bars, cells deprived of either leucine or histidine to which the deprived amino acid was returned; gray bars, cells maintained in serum-free DMEM. B, the guanine nucleotide exchange activity of eIF2B was measured in extracts of L6 cells as described under "Experimental Procedures." The results represent the mean ϮS.E. of 11-13 experiments/condition. Within each experiment 2-3 cultures/condition were individually analyzed. C, the amount of phosphorylated eIF2␣ present in homogenates of L6 myoblasts was determined by protein immunoblot analysis using an affinity-purified antibody specific for eIF2␣ phosphorylated on Ser 51 as described under "Experimental Procedures." The results represent the mean ϮS.E. of four experiments. Within each experiment 2-3 cultures/condition were individually analyzed. *, p Ͻ 0.001 versus leucine-deprived cells to which leucine was returned; †, p Ͻ 0.001 versus histidine-deprived cells to which histidine was returned; **, p Ͻ 0.05 versus leucine-deprived cells to which leucine was returned; † †, p Ͻ 0.05 versus control cells.

FIG. 2.
Effects of leucine or histidine deprivation and resupplementation on the phosphorylation of eIF2␣ in L6 myoblasts. L6 myoblasts were maintained in culture as described in the legend to Fig. 1. Cells were harvested in SDS sample buffer at 90°C, and homogenates were resolved by isoelectric focusing on slab gels as described under "Experimental Procedures." eIF2␣ was quantitated by protein immunoblot analysis as described under "Experimental Procedures." The results shown are representative of three experiments that were performed. The positions of the phosphorylated (eIF2(␣P)) and unphosphorylated (eIF2␣) forms of eIF2␣ are noted to the right of the figure. control, cells maintained in serum-free DMEM; Ϫhis, cells deprived of histidine; his addback, histidine-deprived cells to which histidine was returned; Ϫleu, cells deprived of leucine; leu addback, leucine-deprived cells to which leucine was returned. eIF2B for eIF2(␣P) results in the effective sequestration of eIF2B into an inactive complex, with eIF2(␣P) inhibiting eIF2B on an approximately equimolar basis. Thus, to inhibit eIF2B activity completely, an equimolar amount of eIF2(␣P) must be present. To establish whether the change in eIF2␣ phosphorylation shown in Fig. 1C could account for the observed change in protein synthesis, both the proportion of eIF2␣ in the phosphorylated form as well as the molar concentrations of eIF2 and eIF2B were determined. As shown in Fig. 2, phosphorylated eIF2␣ was almost undetectable in control cells. Furthermore, although the exact proportion in the phosphorylated form was too low to be accurately measured, the proportion that was phosphorylated in amino acid-deprived cells was no more than 5% that of the total. Because the cells were harvested directly into SDS sample buffer at 90°C, it is unlikely that the low proportion of eIF2␣ in the phosphorylated form was a result of dephosphorylation that occurred during sample preparation.
In most eukaryotic cells, the molar ratio of eIF2 to eIF2B is in the range of 2-5:1 (17,27). If the ratio is similar in L6 myoblasts, then an increase in the amount of eIF2(␣P) to 5% that of total eIF2 could not account for the 50% decrease in protein synthesis caused by amino acid deprivation. Therefore, the amounts of the two proteins were measured in L6 myoblasts using protein immunoblot analysis. As shown in Fig. 3, the ratio of eIF2 to eIF2B in L6 myoblasts is approximately 2:1. Thus, if 1 mol of eIF2(␣P) inhibits approximately 1 mol of eIF2B, then phosphorylation of eIF2␣ cannot account completely for the inhibition of protein synthesis caused by amino acid deprivation in L6 myoblasts.
An alternative mechanism for the change in eIF2B activity in response to leucine and histidine availability might be a change in the phosphorylation state of the ⑀-subunit of eIF2B. However, a direct examination of eIF2B⑀ phosphorylation, for example by 32 P i labeling, is complicated because the protein has at least four sites that can be phosphorylated by three known eIF2⑀ kinases (28 -30). In an attempt to circumvent this problem, others (31, 32) have measured eIF2B⑀ kinase activity   FIG. 3. Contents of eIF2 and eIF2B in L6 myoblasts. The contents of eIF2 and eIF2B in L6 myoblasts were determined by protein immunoblot analysis using eIF2␣ and eIF2B⑀ expressed in and purified from Sf9 cells (16,67) as standards. The top and middle panels represent the results from typical immunoblots using monoclonal anti-eIF2␣ and anti-eIF2B⑀ antibodies, respectively. The eIF2B⑀ antibody was raised against eIF2B that was purified from rat liver as described previously (17). The molecular mass of the proteins used as standards is greater than the corresponding proteins in L6 cells, because the eIF2␣ and eIF2B⑀ were expressed in Sf9 cells with an 8-amino acid extension at the N terminus to aid in purification. The results shown in the bottom panel represent the mean ϮS.E. of samples from 18 dishes of cells from 3 different experiments and are presented relative to total cell protein.

FIG. 4. Effects of leucine or histidine deprivation on GSK-3 and eIF2B⑀ kinase activities.
L6 myoblasts were maintained in culture as described in the legend to Fig. 1. A, GSK-3 activity was measured as described under "Experimental Procedures" and is presented as the rate of incorporation of 32 P i into a peptide substrate. The results represent the mean ϮS.E. of samples for three dishes of cells for each condition. B, eIF2B⑀ kinase activity was measured in L6 myoblast extracts as described under "Experimental Procedures." Results from a representative assay are shown as an inset to the figure. eIF2B⑀ kinase activity is presented as the rate of incorporation of 32 P i into eIF2B⑀. The results represent the mean ϮS.E. of samples from six dishes of cells for each condition. *, p Ͻ 0.05 versus cells maintained in complete medium. rather than 32 P i incorporation. In those studies, GSK-3 was implicated as the kinase that mediated the increase in eIF2B activity in insulin-stimulated CHO.T cells. However, in the present study, no change in GSK-3 activity was observed in extracts of L6 myoblasts deprived of either leucine or histidine (Fig. 4A). Because kinases other than GSK-3 have been shown to phosphorylate eIF2B⑀, we also measured eIF2B⑀ kinase activity in extracts of L6 myoblasts using purified, recombinant eIF2B⑀ as substrate. As shown in Fig. 4B, deprivation of either leucine or histidine caused a significant decrease in eIF2B⑀ kinase activity. Therefore, changes in eIF2B activity in response to leucine or histidine deprivation correlate directly with changes in eIF2B⑀ kinase activity, and the kinase activity involved is unlikely to be GSK-3.
To further define the mechanisms through which amino acids regulate translation initiation, the effect of leucine or histidine deprivation on a second potential regulatory step, the binding of mRNA to the 40 S ribosomal subunit, was examined. Under a variety of conditions, the binding of mRNA to the ribosome is regulated through modulation of the association of eIF4E with eIF4G and 4E-BP1 (33,34). In the present study, the effect of leucine or histidine deprivation on the association of eIF4E with eIF4G was examined by immunoprecipitation of eIF4E and the eIF4E⅐eIF4G complex from L6 myoblast homogenates followed by protein immunoblot analysis to quantify the amount of the two proteins in the immunoprecipitate. As shown in Fig. 5A, leucine deprivation caused a 50% decrease in the amount of eIF4G that co-precipitated with eIF4E. The readdition of leucine to the medium of leucine-deprived cells rapidly returned the amount of eIF4G associated with eIF4E to control values. In contrast to the results observed with leucine, deprivation of histidine had no effect on the association of eIF4E with eIF4G, despite the fact that histidine deprivation was equally effective as leucine in inhibiting protein synthesis.
The association of eIF4E with 4E-BP1 was similarly examined by immunoprecipitation of the 4E-BP1⅐eIF4E complex using a monoclonal anti-eIF4E antibody. The amount of 4E-BP1 associated with eIF4E was approximately 5-fold greater in immunoprecipitates from leucine-deprived cells compared with control cells (Fig. 5B). In agreement with previous studies (5,35,36), only the ␣and ␤-forms of 4E-BP1 were found associated with eIF4E. Readdition of leucine to the medium of leucine-deprived cells decreased the amount of 4E-BP1 bound to eIF4E to a value not significantly different from the control. As expected based on the lack of change in eIF4G binding to eIF4E, histidine deprivation had no effect on the amount of 4E-BP1 present in the immunoprecipitate.
The only known mechanism for regulating the association of eIF4E and 4E-BP1 involves modulation of the phosphorylation state of 4E-BP1 (33). The phosphorylation state of 4E-BP1 can conveniently be examined by resolution of the phosphorylated forms of the protein during SDS-polyacrylamide gel electrophoresis. During SDS-polyacrylamide gel electrophoresis, 4E-BP1 is resolved into multiple electrophoretic forms, termed ␣, ␤, and ␥, representing differentially phosphorylated forms of the protein. The most highly phosphorylated form, i.e. the ␥-form, exhibits the slowest electrophoretic mobility and is the only one of the three that does not bind to eIF4E. Therefore, in the present study, phosphorylation of 4E-BP1 is expressed as the percentage of the protein in the ␥-form. As shown in Fig.  5C, leucine deprivation resulted in an approximate 4-fold decrease in the amount of 4E-BP1 in the ␥-form, and leucine readdition to the medium of leucine-deprived cells restored 4E-BP1 phosphorylation to the control value. Alterations in histidine availability had no effect on 4E-BP1 phosphorylation.
The regulation of 4E-BP1 phosphorylation by hormones such as insulin (13,36) or insulin-like growth factor-1 (35) occurs through activation of a signal transduction pathway that also results in activation of p70 S6k . Activation of p70 S6k is associated with its phosphorylation on multiple Ser and Thr residues. Upon activation, p70 S6k typically resolves into multiple electrophoretic forms after separation by electrophoresis on SDS-polyacrylamide gels, with increased phosphorylation being associated with decreased electrophoretic mobility (37). Therefore, in the present study, the effect of leucine or histidine deprivation on the phosphorylation of p70 S6k was investigated in L6 myoblast homogenates by protein immunoblot analysis. As shown in Fig. 6A, deprivation of leucine caused a shift in the distribution of p70 S6k to the faster migrating species, indicative of dephosphorylation of the protein. After addition of leucine to leucine-deprived cells, the distribution of the kinase among the electrophoretic forms returned to the pattern observed in control cells. This observation suggests that an amino acid, leucine, can alter the phosphorylation of a kinase that is part of an important signal transduction pathway. In contrast, histidine deprivation had no effect on p70 S6k phosphorylation.
To further define the importance of the changes in association of eIF4E with eIF4G and 4E-BP1 in the regulation of global protein synthesis in L6 myoblasts, the above experiments were repeated in the presence of insulin. A number of studies (13,36,38,39) have shown that insulin activates the p70 S6k signal transduction pathway and causes phosphorylation of 4E-BP1. If leucine and insulin stimulate 4E-BP1 and p70 S6k phosphorylation through the same signal transduction pathway, it might be expected that insulin would prevent the dephosphorylation of the two proteins caused by leucine depri-vation. As shown in Fig. 6B, insulin caused a decrease in the electrophoretic mobility of p70 S6k , indicative of phosphorylation of the protein. Leucine deprivation caused a slight increase in the electrophoretic mobility of p70 S6k in the presence of insulin, although this result was only observed in about 50% of the experiments. In the remainder of the experiments, the insulin-stimulated phosphorylation of p70 S6k was unaffected by leucine deprivation. However, even in those experiments where leucine deprivation caused partial dephosphorylation of p70 S6k in the presence of insulin, p70 S6k was still more highly phosphorylated than in control cells in the absence of the hormone. Thus, insulin prevented the dephosphorylation of p70 S6k caused by leucine deprivation. As was observed in the absence of insulin, histidine deprivation had no effect on p70 S6k phosphorylation.
The leucine-and histidine-induced alterations in protein synthesis observed in the absence of insulin were maintained in the presence of the hormone (Fig. 7A). Thus, deprivation of either leucine or histidine reduced protein synthesis to approximately 45% that of the value observed in cells maintained in complete medium without insulin. Addition of leucine or histidine to the medium of cells deprived of the corresponding amino acid resulted in a restoration of protein synthesis to control values. Under the conditions used in this study, insulin had no significant effect on the overall rate of protein synthesis in cells maintained in complete, serum-free medium. The observation that insulin did not stimulate protein synthesis under these conditions is discussed later.
In agreement with the lack of effect of insulin on the overall rate of protein synthesis, the hormone had no effect on the amount of eIF4G bound to eIF4E in cells maintained in complete medium (Fig. 7B). However, insulin completely prevented the decrease in eIF4G associated with eIF4E caused by leucine deprivation. Insulin also prevented the increased association of 4E-BP1 with eIF4E (Fig. 7C) as well as the dephosphorylation of 4E-BP1 (Fig. 7D) caused by leucine deprivation. In fact, even in cells maintained in complete medium, insulin caused a decrease in the binding of 4E-BP1 to eIF4E as well as stimulated the phosphorylation of 4E-BP1 (compare Figs. 5B with 7C and Figs. 5C with 7D, respectively). The results demonstrate that the leucine-induced alterations in the association of eIF4E with eIF4G and 4E-BP1 are not rate-controlling for global protein synthesis under these conditions. DISCUSSION The translation of mRNA into protein has most often been examined at a global level, i.e. regulation involving synthesis of all, or almost all, proteins (reviewed in Ref. 40). One established mechanism through which cells regulate global protein synthesis is through changes in the activity of eIF2B. Translation of all proteins is dependent upon the binding of the eIF2⅐GTP⅐tRNA i Met ternary complex to the 40 S ribosomal subunit, and the amount of eIF2 present in the ternary complex is dependent upon the activity of eIF2B (2,3). Furthermore, because translation of greater than 90% of the cytoplasmic mRNAs in mammalian cells occurs through a cap-dependent process involving eIF4E, it might be expected that modulation of the availability of eIF4E would result in changes in global protein synthesis. Indeed, changes in the distribution of eIF4E between the active eIF4E⅐eIF4G and the inactive eIF4E⅐4E-BP1 complexes have been suggested as the mechanism through which hormones such as insulin or insulin-like growth factor-1 regulate global protein synthesis (13,35,36,39). However, it is noteworthy that protein synthesis has not been measured in the majority of the studies examining the effect of hormones on the association of eIF4E and 4E-BP1 (6,35,39,(41)(42)(43).
In contrast to studies linking eIF4E to the regulation of global protein synthesis, other studies using genetic methods to alter the expression of eIF4E suggest that changes in the cellular amount of eIF4E result in the synthesis of select proteins. For example, in NIH3T3 cells overexpressing eIF4E, a specific increase in the synthesis of cyclin D1 and ornithine decarboxylase occurs (44 -47). Furthermore, overexpression of eIF4E results in transformation of NIH3T3 cells, an effect that is reversed by coexpression of either 4E-BP1 or 4E-BP2 (48). These results suggest that the availability of eIF4E per se is important in transformation and regulation of translation of specific mRNAs.
In the present study, the mechanism through which leucine and histidine regulate global protein synthesis was examined. It was found that leucine and histidine both modulated global protein synthesis and eIF2B activity, whereas leucine, but not histidine, modulated the availability of eIF4E. Furthermore, the changes in eIF4E distribution and 4E-BP1 phosphorylation were not observed in the presence of insulin even though alterations in protein synthesis still occurred. Thus, in L6 myoblasts deprived of amino acids, changes in the amount of eIF4E bound to eIF4G were not related to the observed changes in global protein synthesis. Instead, the changes in eIF4E distribution were consistent with the possibility that leucine modulates the translation of specific mRNAs through changes in the availability of eIF4E.
The leucine-induced changes in eIF4E distribution observed in the present study were directly related to alterations in the phosphorylation state of 4E-BP1. Previous studies showed that hormones such as insulin (13,39,49), insulin-like growth factor-1 (35), or platelet-derived growth factor (35) stimulate phosphorylation of 4E-BP1 through the p70 S6k pathway. Use of inhibitors of phosphatidylinositol 3-kinase and the mammalian target of rapamycin (mTOR) have further implicated the p70 S6k signaling pathway in the phosphorylation of 4E-BP1 (13,35,38,39,50). A more recent study demonstrated that in rat brain extracts, mTOR directly phosphorylates 4E-BP1 and that phosphorylation by mTOR results in a decrease in the affinity of eIF4E for 4E-BP1 (51). In addition, Burnett et al. (52) reported that mTOR phosphorylates both 4E-BP1 and p70 S6k . mTOR is a member of the p70 S6k signaling pathway and is currently thought to be downstream of phosphatidylinositol 3-kinase but upstream of p70 S6k in the pathway (53). In the present study, leucine stimulated the phosphorylation of both p70 S6k and 4E-BP1, consistent with the hypothesis that leucine regulated phosphorylation of 4E-BP1 through activation of the p70 S6k signaling pathway. In support of this idea, rapamycin attenuated the leucine-induced phosphorylation of p70 S6k in L6 myoblasts. 2 Furthermore, studies by others (54,55) showed that amino acids stimulate phosphorylation of ribosomal protein S6 in rat hepatocytes and that rapamycin prevents the stimulation. The results suggest that leucine regulates the phosphorylation of 4E-BP1 through a common signal transduction pathway utilized by insulin and other hormones. However, a recent study (56) showing that leucine does not activate phosphatidylinositol 3-kinase in Fao cells suggests that the mechanism through which leucine activates the p70 S6k pathway is distinct from that used by insulin to activate the pathway.
In contrast to the differential effect of leucine and histidine on the distribution of eIF4E, both amino acids modulated the activity of eIF2B. Furthermore, changes in eIF2B activity correlated positively with alterations in global protein synthesis both in the presence (data not shown) and in the absence of insulin. Thus, the changes in global protein synthesis appear to be because of modulation of eIF2B activity rather than through the availability of eIF4E. An inhibition of global protein synthesis also occurs in a variety of mammalian cells exposed to elevated temperatures. The inhibition that occurs in response to heat shock has been attributed to either inhibition of eIF2B (57) or eIF4E (58,59). A more recent study where both proteins were examined (60) reached a similar conclusion to that reported herein. In that study, heat shock caused both an inhibition of eIF2B activity as well as alterations in the phosphorylation of 4E-BP1 and association of eIF4E with 4E-BP1. However, in some cell types, e.g. H35-Reuber hepatoma cells, heat shock caused a decrease in association of eIF4E with 4E-BP1, whereas in other cell types, e.g. CHO cells, heat shock caused the exact opposite effect. The results led those authors to propose that global protein synthesis was regulated during heat shock by eIF2B and that phosphorylation of 4E-BP1, with the concomitant changes in association of eIF4E with 4E-BP1, might influence translation of specific mRNAs.
We (13) and others (61) reported that insulin stimulates protein synthesis in L6 myoblasts. In addition, insulin enhances the activity of eIF2B in both skeletal muscle (62,63) and CHO cells overexpressing the human insulin receptor (CHO.IR or CHO.T cells) (64). In contrast, in the present study insulin had no effect on either protein synthesis or eIF2B activity in L6 myoblasts. The basis for this apparent discrepancy lies in the length of time during which the cells were deprived of serum. In our earlier study (13), insulin only stimulated protein synthesis in L6 myoblasts that had been deprived of serum for at least 18 h. A similar phenomenon has been reported for chick embryo fibroblasts (65). Thus, insulin does not stimulate protein synthesis above the rate observed in cells maintained in serum-containing medium. Rather, the stimulatory effect of insulin is observed only after protein synthesis becomes impaired in response to serum deprivation. In the present study, L6 myoblasts were deprived of serum for a total of only 2 h, and the rate of protein synthesis at that time was the same in serum-deprived as in serum-containing medium (data not shown).
The best characterized mechanism for regulating the activity of eIF2B is through phosphorylation of the ␣-subunit of eIF2 (reviewed in Ref. 2). Phosphorylation of eIF2␣ converts eIF2 from a substrate into a competitive inhibitor and results in the effective sequestration of eIF2B into an inactive complex. In the present study, an increase in eIF2␣ phosphorylation in response to deprivation of either leucine or histidine is likely to account at least partially for the observed changes in eIF2B activity. However, because the proportion of eIF2␣ in the phosphorylated form was less than 5% that of the total protein and the molar ratio of eIF2 to eIF2B was approximately 2:1, it is unlikely that phosphorylation of eIF2 accounts completely for the inhibition of eIF2B activity or protein synthesis. An alternative explanation for the inhibition of eIF2B activity may be a change in the phosphorylation state of the ⑀-subunit of the protein. Three kinases have been identified that phosphorylate eIF2B⑀ in vitro; GSK-3, casein kinase-I, and casein kinase-II (29,31,66). In CHO.T cells, phosphorylation of eIF2B⑀ by GSK-3 is an important mechanism for regulation of the guanine nucleotide exchange activity of eIF2B (64). In those cells, insulinregulatesGSK-3throughaphosphatidylinositol3-kinasedependent pathway (64). However, Patti et al. (56) reported that leucine does not activate phosphatidylinositol 3-kinase in Fao cells. Furthermore, in the present study, we show that deprivation of either leucine or histidine has no effect on GSK-3 activity in L6 myoblasts. Finally, the direction of change in eIF2B⑀ kinase activity observed in leucine-or histidine-deprived L6 myoblasts is the opposite of what would be expected if the eIF2B⑀ kinase were GSK-3. Thus, the stimulation of eIF2B activity by insulin in CHO.T cells is associated with an inhibition of GSK-3, suggesting that GSK-3 inhibits eIF2B (31,32). Overall, it is unlikely that GSK-3 mediates the regulation of eIF2B activity by amino acids. However, because eIF2B⑀ kinase activity decreases in response to amino acid deprivation, a role for other kinases in the regulation of eIF2B activity in L6 myoblasts must be considered.