Insulin Signaling and the General Amino Acid Control Response

ATF4 is a transcription factor that induces a genetic program for amino acid synthesis and amino acid uptake. Previous work demonstrated that ATF4 expression is increased either by insulin or by the general amino acid control (GAAC) response, an evolutionarily ancient pathway that is activated when eukaryotic cells are deprived of amino acids. It is not known whether insulin and the GAAC pathway increase ATF4 expression by the same or different mechanisms. In these studies, we demonstrate that insulin-mediated ATF4 expression occurs as part of a coordinated anabolic program that does not require an essential component of the GAAC pathway, the protein kinase GCN2. Moreover, insulin and the GAAC pathway have an additive effect on expression of ATF4 and downstream mRNAs for amino acid synthesis and uptake. These data suggest that the GAAC pathway may facilitate insulin-mediated anabolism when exogenous amino acids are limiting. We conclude that insulin signaling and the GAAC response comprise two distinct yet complimentary pathways to ATF4 expression, allowing anabolism to be finely tuned to amino acid availability.

ATF4 is a transcription factor that induces a genetic program for amino acid synthesis and amino acid uptake. Previous work demonstrated that ATF4 expression is increased either by insulin or by the general amino acid control (GAAC) response, an evolutionarily ancient pathway that is activated when eukaryotic cells are deprived of amino acids. It is not known whether insulin and the GAAC pathway increase ATF4 expression by the same or different mechanisms. In these studies, we demonstrate that insulin-mediated ATF4 expression occurs as part of a coordinated anabolic program that does not require an essential component of the GAAC pathway, the protein kinase GCN2. Moreover, insulin and the GAAC pathway have an additive effect on expression of ATF4 and downstream mRNAs for amino acid synthesis and uptake. These data suggest that the GAAC pathway may facilitate insulin-mediated anabolism when exogenous amino acids are limiting. We conclude that insulin signaling and the GAAC response comprise two distinct yet complimentary pathways to ATF4 expression, allowing anabolism to be finely tuned to amino acid availability.
In mammals, changes in nutrient availability induce changes in levels of metabolic hormones, which in turn orchestrate the metabolism of trillions of cells, allowing them to cooperate and function as a single organism (1,2). Conversely, in unicellular organisms, metabolism is largely governed by nutrient-sensing pathways, which sense changes in nutrient availability and generate an adaptive, cell autonomous metabolic response (3). Interestingly, many nutrient-sensing pathways of unicellular eukaryotes have been retained in higher organisms such as mammals; one such nutrient-sensing pathway that is highly conserved is the general amino acid control (GAAC) 2 pathway, which senses amino acid deficiencies and restores intracellular amino acid levels (4). The GAAC pathway was originally described in yeast and was later found to be present in mammalian cells (4 -11).
Several molecular aspects of the GAAC pathway are well defined (4); amino acid deprivation increases levels of uncharged tRNAs, which bind to and activate the kinase GCN2. GCN2 phosphorylates the eukaryotic translation initiation factor 2␣ (eIF2␣), thereby inhibiting global protein synthesis but increasing synthesis of a few specific proteins. One such protein that is increased under these conditions is a basic leucine zipper transcription factor that activates a large panel of genes encoding amino acid transporters and amino acid biosynthetic enzymes. In yeast, this transcription factor is referred to as GCN4 (4); the mammalian counterpart is activating transcription factor 4 (ATF4, also known as CREB2 (5)(6)(7)(8)(9)(10)(11)). In addition to stimulating ATF4 mRNA translation, amino acid deprivation increases levels of GCN4 or ATF4 mRNA (4,6). The end result of the GAAC pathway is activation of genes that correct amino acid deficiencies and restore homeostasis.
ATF4 is also subject to opposing regulation by hormones that play a central role in regulating mammalian metabolism (12). Glucocorticoids (catabolic hormones that are required for mammalian survival during prolonged fasting or stress) decrease levels of ATF4 mRNA and protein, leading to the repression of ATF4-dependent mRNAs encoding amino acid transporters and amino acid biosynthetic enzymes. Glucocorticoid-mediated repression of ATF4 is coupled with inhibition of global mRNA translation and protein synthesis and the activation of genes for protein catabolism. Conversely, insulin (an anabolic hormone) overcomes all of these effects of glucocorticoids. Thus, insulin increases levels of ATF4 mRNA, ATF4 protein, ATF4-dependent mRNAs, and global protein synthesis, and insulin represses genes for protein catabolism. Taken together, these data suggest that ATF4 mediates a portion of the coordinated anabolic response to insulin and that ATF4 is repressed as part of the coordinated catabolic response to glucocorticoids. Indeed, the in vivo role of ATF4 in mammalian anabolism has been established through studies of ATF4-deficient mice, which exhibit pleiotropic growth defects (13)(14)(15)(16), at least some of which are rescued by a high protein diet (17).
These considerations led us to ask the following question. What is the relationship between evolutionarily ancient nutrient-sensing pathways and the hormonal signaling pathways of multicellular organisms? With regards to ATF4, one possibility is that glucocorticoids and insulin regulate the GAAC pathway to control ATF4 and genes needed for amino acid synthesis and uptake. For example, insulin-mediated protein synthesis might induce a state of amino acid depletion within cells, thereby acti-vating GCN2 and increasing ATF4; or insulin could directly activate GCN2 through a previously unrecognized mechanism. Alternatively, the integrated mammalian responses to fasting and feeding may have evolved in such a way that glucocorticoids and insulin regulate ATF4 independently of the GAAC pathway. Our goal in these studies was to address this question.
Cell Culture-All studies described here utilized mouse L cells (D9 strain (18)). Tissue culture medium A is Dulbecco's modified Eagle's medium containing 1 g/liter of glucose (Mediatech catalog number 10-014) containing antibiotics (100 units/ml penicillin, 100 g/ml streptomycin sulfate) and 10% (v/v) fetal bovine serum; medium B is medium A lacking antibiotics; medium C is medium A lacking fetal bovine serum. Cells were maintained in monolayer in medium A at 37°C and an atmosphere of 8 -9% CO 2 . With the exception of experiments using RNA interference (as described below), cells were set up for experiments in medium A on day 0 at a density of 7 ϫ 10 5 cells/100-mm dish (for use on day 2) or 3.5 ϫ 10 5 cells/100-mm dish (for use on day 3). Prior to the incubation protocols described in the figure legends, the cells were washed twice with phosphate-buffered saline (PBS).
RNA Interference-Mouse L cells were set up at a density of 2.5 ϫ 10 5 cells/60-mm dish on day 0. On day 1, cells were washed with PBS, refed with 3 ml of medium B, followed by the addition of a mixture containing 1 ml of OptiMEM (Invitrogen), 10 l of Lipofectamine 2000 reagent (Invitrogen) and 200 nM siRNA duplex (to give a final siRNA concentration of 50 nM). Control cells were transfected with siCONTROL nontargeting siRNA (Dharmacon). The siRNA duplex targeting GCN2 (also known as eukaryotic translation initiation factor 2␣ kinase 4 (Eif2ak4); GenBank TM accession number NM_013719) contained the following nucleotide sequence (5Ј to 3Ј): sense GGAAAUCA-CUAGCUUGACAUU; antisense UGUCAAGCUAGU-GAUUUCCUU. Similar results were obtained using an individual siRNA duplex targeting a different region of the GCN2 mRNA (containing the following nucleotide sequence (5Ј to 3Ј): sense GCGCUGCGUUCUUCAGUGAUU; antisense UCA-CUGAAGAACGCAGCGCUU) or using a mixture of four siRNA duplexes targeting GCN2 (Dharmacon catalog number D-044353-02) (not shown). On day 2 (24 h after transfection), cells were re-fed with serum-free medium C and then subjected to varying conditions of incubation as described in the legends for Figs. 2 and 5. Following incubation, the cells from three identically treated dishes were pooled and harvested for total cellular RNA followed by analysis using quantitative real-time PCR (qPCR).
Analysis of RNA Levels by qPCR-Total cellular mRNA was isolated and treated with DNase I using the RNeasy kit (Qiagen), according to the manufacturer's protocol. Small RNA (for the analysis of pre-tRNAs) was isolated using the Mirvana kit (Ambion), according to the manufacturer's small RNA enrich-

mRNA EXPRESSION IN DEX-TREATED CELLS
. Histidinol overcomes glucocorticoid-mediated repression of ATF4 and ATF4-dependent mRNAs. A, incubation protocol to examine the effects of dexamethasone (Dex), insulin, and histidinol. B, mouse L cells were preincubated for 48 h in medium C in the absence or presence of dexamethasone (10 nM) and then for an additional 6 h in the absence or presence of insulin (100 nM) or histidinol (2 mM), as indicated. In this and all other experiments, histidinol or insulin was added directly to the existing cell culture medium (in the continued presence of dexamethasone). Following incubation, nuclear protein extracts were analyzed by immunoblot using the anti-ATF4 polyclonal antibody. The data shown are representative of three independent experiments. C, mouse L cells were preincubated for 48 h in medium C containing dexamethasone (10 nM) and then for an additional 6 h in the absence or presence of insulin (100 nM) or histidinol (2 mM), as indicated. Following incubation, total cellular RNA was analyzed by qPCR, and transcript levels were normalized to the transcript levels in cells incubated in the absence of histidinol and insulin, which were set at 1. The representative mRNAs involved in amino acid synthesis or uptake encode, from right to left: ATF4; asparagine synthetase (Asns); phosphoserine aminotransferase 1 (Psat1); methylenetetrahydrofolate dehydrogenase (NAD ϩdependent), methenyltetrahydrofolate cyclohydrolase (Mthfd2); the Slc1a4, solute carrier family 1, member 4 (also known as ASCT1) (Slc1a4); the solute carrier family 7, member 1 (also known as CAT-1) (Slc7a1); and the solute carrier family 7, member 5 (also known as LAT-1) (Slc7a5). The representative mRNAs involved in lipid uptake or synthesis encode, from right to left: the low density lipoprotein receptor (LDL-R); HMG-CoA reductase (HMGCoAR); acetyl-CoA carboxylase (ACC); and fatty acid synthase (FAS). The representative mRNAs involved in protein or lipid catabolism encode, from right to left: atrogin-1 and pyruvate dehydrogenase kinase, isoenzyme 4 (PDK4). B and C, the data shown are representative of three independent experiments. ment protocol. First strand cDNA was synthesized from 2 g of RNA using random hexamer primers and the TaqMan reverse transcription kit (catalog number N808-0234; Applied Biosystems). Specific primers for each gene were designed using the Primer Express software (Applied Biosystems). The nucleotide sequence of the GCN2 primers were: 5Ј-CGTTTCTCAGC-GAGCATAACAA-3Ј and 5Ј-CCTGAGCCTGCCTTTCCA-3Ј. All of the other qPCR primer sequences have been described previously (12, 19 -21). The real-time PCR contained, in a final volume of 20 l, 2 ng of reverse-transcribed RNA, 167 nM forward and reverse primers, and 10 l of Power SYBR Green PCR master mix (catalog number 4367659; Applied Biosystems). PCR was carried out using the 7500 Fast real-time PCR system (Applied Biosystems). All qPCR reactions were performed in triplicate, and the cycle threshold (Ct) values were averaged to give the final results. The relative amounts of all RNAs were calculated using the comparative Ct method (22). mRNA encoding 36B4 was used as the invariant control in all studies.
Metabolic Labeling-[ 3 H]Leucine (120 Ci/mmol; 20 Ci/ dish) was added directly to the cell culture medium, after which one dish from each experimental condition was placed at 4°C (to measure background incorporation of [ 3 H]leucine), whereas three dishes from each condition were placed back in the 37°C incubator. Following incubation, cells were washed twice with PBS and then harvested, and an aliquot of cells was subjected to precipitation using 10% (w/v) trichloroacetic acid in the presence of 0.45% salmon sperm DNA. Trichloroacetic acid precipitates were then applied to glass fiber filters set upon a vacuum manifold. Filters were sequentially washed with 10% trichloroacetic acid, 5% trichloroacetic acid, and 95% ethanol and then placed in scintillation mixture for measurement of acid-insoluble radioactivity. The final results were obtained by averaging the counts obtained from the three dishes in each condition that were incubated at 37°C and then subtracting the background count obtained from the dish incubated at 4°C and normalizing the counts to the total mg of protein/dish under each condition.
Preparation of Nuclear Extracts, SDS-PAGE, and Immunoblot Analysis-Because ATF4 has an extremely short half-life (30 -60 min) secondary to proteasomal degradation (23,24), the cells were treated with the proteasome inhibitor N-acetylleucinal-leucinal-norleucinal (25 g/ml) for the final 1 h of incubation, as is routinely done in studies of proteins with rapid turnover (12,25). Following incubation, the cells from three identically treated 100-mm dishes of cells were washed once with cold PBS, scraped and pooled into 50-ml tubes on ice, and then centrifuged at 1000 ϫ g for 5 min. Cell pellets were resuspended in 1.2 ml of buffer A and then passed through a 22-gauge needle 22 times followed by centrifugation at 1000 ϫ g for 5 min. The 1000 ϫ g pellet was resuspended in 250 l of buffer C, rotated at 4°C for 1 h, and then centrifuged at 20,000 ϫ g for 20 min to remove insoluble material. The protein concentration was then measured using the BCA kit (Pierce), after which a 25-g aliquot was mixed with 0.25 volume of buffer B and heated for 5 min at 95°C. Samples were subjected to 10% SDS-PAGE and then transferred to Hybond-C extra nitrocellulose filters (Millipore). Immunoblots were performed at room temperature using a 1:1000 dilution of a rabbit poly-clonal antiserum SC-200 (Santa Cruz Biotechnology), which recognizes the COOH terminus of mouse ATF4.
Measurement of Intracellular Free Amino Acids-Following incubation, cells were washed three times with cold PBS and scraped into 10% (v/v) methanol (150 l/60-mm dish). Lysates from two identically treated dishes of cells were pooled, protein concentrations were determined using the BCA method, and free (nonhydrolyzed) amino acid concentrations were determined using cation exchange chromatography (Hitachi L-8800 amino acid analyzer), as described previously (12).

Activation of the GAAC Pathway by Histidinol Overcomes
Glucocorticoid-mediated Repression of ATF4-Mouse L fibroblasts are an immortalized cell line with a previously defined response to glucocorticoids and insulin (12,26). When mouse L cells are incubated in serum-free medium containing glucocorticoids, they enter a metabolically quiescent state where expression of ATF4 and ATF4 target genes is repressed. Glucocorticoid-mediated repression of ATF4 is coupled with repression of . Total cellular RNA was analyzed by qPCR, and transcript levels in the presence of histidinol were normalized to transcript levels in cells incubated in the absence of histidinol and harvested at the same time point, which were set at 1. Asns, asparagine synthetase; Slc1a4, the Slc1a4, solute carrier family 1, member 4 (also known as ASCT1); Slc7a5, the solute carrier family 7, member 5 (also known as LAT-1). B, following preincubation, cells were incubated for 6 h in the indicated concentration of histidinol. Total cellular RNA was analyzed by qPCR, and transcript levels were normalized to transcript levels in cells incubated in the absence of histidinol, which were set at 1. C, mouse L cells were transfected with the indicated siRNA duplex and then switched to medium C containing dexamethasone (10 nM) for 48 h. Cells were then incubated for an additional 6 h in the absence or presence of histidinol (2 mM). Total cellular RNA was analyzed by qPCR, and transcript levels were normalized to transcript levels in cells transfected with the same siRNA but incubated in the absence of histidinol, which were arbitrarily set at 1. Data represent the average values obtained from four independent experiments, and error bars represent S.E. The GCN2 siRNA specifically reduced basal levels of GCN2 mRNA by 88%; histidinol did not alter GCN2 mRNA levels (not shown). LDL-R, the low density lipoprotein receptor; HMGCoAR, HMG-CoA reductase; Slc7a1, the solute carrier family 7, member 1 (also known as CAT-1).
ATF4-independent anabolic genes for lipid synthesis and uptake and the induction of catabolic genes for protein and lipid breakdown (12). Under these conditions, a physiologic concentration of insulin dominantly overrides the effects of glucocorticoids, thus repressing catabolic genes and inducing expression of ATF4 and both ATF4-dependent and ATF4-independent anabolic genes (12).
To test the hypothesis that activation of the GAAC pathway might overcome the repressive effect of glucocorticoids on ATF4, we used histidinol, an amino alcohol that competitively inhibits histidinyl-tRNA synthetase, thereby increasing levels of uncharged tRNA His and activating GCN2 (7,27). Our incubation protocol is shown in Fig. 1A; mouse L cells were incubated in the absence or presence of dexamethasone for 48 h followed by an additional 6-h incubation in the absence or presence of either insulin or histidinol, which was directly added to the cell culture medium in the continued presence of dexamethasone.
In the absence of dexamethasone, ATF4 was expressed at a high level (Fig. 1B, lane 1), and neither insulin (12) nor histidinol (supplemental Fig. 1) increased levels of ATF4 or ATF4-dependent mRNAs. Dexamethasone reduced expression of ATF4 (Fig. 1B, lane 2), and under these conditions, histidinol, like insulin, increased ATF4 protein levels (Fig. 1B, lanes 3 and 4). In contrast to insulin (Fig. 1C, left side), histidinol did not increase levels of mRNAs involved in lipid synthesis and uptake nor decrease levels of mRNAs involved in protein or lipid breakdown (Fig. 1C,  right side). Rather, histidinol specifically increased the mRNA encoding ATF4, which was followed by an increase in ATF4-dependent mRNAs encoding amino acid biosynthetic enzymes and amino acid transporters (Fig. 2A). The effect of histidinol on ATF4 and ATF4-dependent mRNAs was maximal at a concentration of 2 mM (Fig. 2B) and was inhibited by siRNA-mediated silencing of GCN2 (Fig. 2C). These data indicate that the GAAC pathway remains intact and functional in the presence of glucocorticoids.
Glucocorticoids Coordinately Reduce Intracellular Free Amino Acids and tRNA Synthesis-ATF4 is required to maintain levels of intracellular free amino acids (12). Consistent with its repressive effect on ATF4, dexamethasone also reduced levels of intracellular amino acids (Fig. 3A). This presented a paradox. If the GAAC pathway is intact in glucocorticoid-treated cells (Figs. 1-2) and glucocorticoids decrease free intracellular amino acid levels (Fig. 3A), then how do glucocorticoids decrease ATF4? We hypothesized that glucocorticoids might also repress tRNA levels. To test this, we measured levels of several pre-tRNAs (unspliced tRNA precursors that are used as markers of RNA polymerase III activity (20 -21)). Fig. 3B shows that dexamethasone decreased levels of three different pre-tRNAs, which correlated with a reduction in global protein synthesis (Fig. 3C). These data suggest that glucocorticoids uncouple ATF4 from the GAAC pathway by coordinately regulating amino acid and tRNA levels.  ). A, following incubation, free amino acids in cell lysates were measured using ion exchange chromatography, normalized to the total protein concentration in each sample, and then normalized to the levels in cells incubated in the absence of insulin and histidinol, which were set at 1. The data shown represent the sum of all measured species of amino acids. Insulin increased the pooled levels of essential amino acids by 37% and nonessential amino acids by 18%; histidinol increased levels of essential amino acids by 18% and nonessential amino acids by 18% (p Յ 0.03). Levels of individual species of amino acids are found in supplemental Fig. 2. B, following incubation, pre-tRNA levels were analyzed by qPCR and normalized to the pre-tRNA levels in cells incubated in the absence of insulin and histidinol, which were set at 1. C, [ 3 H]leucine was present for the final 2 h of incubation. Following incubation, the amount of [ 3 H]leucine contained in acid-insoluble cellular fractions was measured, normalized to the total protein concentration in each sample, and then normalized to the level in cells incubated in the absence of insulin and histidinol, which was set at 1. A-C, data represent the average values obtained from at least three independent experiments. Error bars represent S.E., and asterisks indicate a significant change (p Ͻ 0.02 by unpaired, two-tailed t test) in cells treated with insulin or histidinol.
Insulin Coordinately Increases Intracellular Amino Acids and tRNA Synthesis-Insulin opposes the repressive effects of glucocorticoids on protein synthesis (2) and on ATF4 and ATF4dependent mRNAs (12) (Fig. 1). Thus, we hypothesized that insulin might also overcome the repressive effect of glucocorticoids on levels of intracellular amino acids and pre-tRNAs. Fig.  4, A-C, shows that insulin coordinately increased levels of free amino acids, pre-tRNAs, and global protein synthesis in dexamethasone-treated cells. In contrast, histidinol increased levels of amino acids (Fig. 4A) but did not increase levels of pre-tRNAs or protein synthesis (Fig. 4, B and C). These data are consistent with the notion that insulin has a generalized anabolic effect that coordinately increases ATF4, amino acids, tRNAs, and protein synthesis, whereas the GAAC pathway increases ATF4 and inhibits global protein synthesis to restore intracellular amino acid levels.
Insulin Increases ATF4 Expression Independently of the GACC Pathway-We reasoned that if insulin and histidinol utilized the same pathway to increase ATF4, then insulin should not increase ATF4 in the presence of a maximal dose of histidinol. On the other hand, if insulin and histidinol increased ATF4 by different mechanisms, then the two compounds might have an additive or synergistic effect. Fig. 5, A and B, shows that insulin and histidinol had an additive effect on ATF4 mRNA levels and synergistic effects on levels of ATF4-dependent mRNAs. Moreover, as shown in Fig. 5C, silencing of GCN2 did not prevent the effect of insulin (top panel) but limited the combined effects of insulin and histidinol on ATF4 mRNA and ATF4-dependent mRNAs (bottom panel). These data suggest that there are two distinct pathways to ATF4 expression: a nutrient-sensing pathway mediated by GCN2 and an anabolic, hormonal signaling pathway that is controlled by insulin.

DISCUSSION
Insulin inhibits several effects of glucocorticoids that comprise an integrated response to starvation or stress (2,28). Thus, insulin has a broad anabolic effect, repressing catabolic genes; inducing genes for lipid anabolism; stimulating global mRNA translation; inducing genes encoding tRNAs; and increasing expression of ATF4. Through ATF4, insulin induces genes for amino acid synthesis and uptake, thereby increasing levels of intracellular amino acids.
ATF4 is also regulated by the GAAC pathway, an evolutionarily ancient nutrient-sensing mechanism found in unicellular organisms that do not rely on hormones such as insulin to regulate metabolism. The GAAC pathway has a more limited effect than the insulin signaling pathway. Like insulin, the GAAC pathway overcomes the repressive effect of glucocorticoids on ATF4 mRNA and protein levels, thereby increasing intracellular amino acids. In contrast to insulin, the GAAC pathway requires GCN2 and does not repress catabolic gene expression, induce genes for lipid anabolism or tRNAs, or stimulate protein synthesis. Indeed, through mechanisms that are ATF4-independent, the GAAC pathway reduces global protein synthesis as part of the adaptive response to amino acid deprivation (4), and interestingly, leads to the repression of hepatic lipogenesis as well (29).
Importantly, the insulin signaling pathway increases ATF4 expression independently of the GAAC pathway. We hypothe-size that these two distinct pathways coexist in mammals so that anabolism can be finely tuned to the amino acid content of the diet (Fig. 6). In our simplified model, mammalian nutrient intake is represented as one of two possibilities: a diet high in amino acids or low in amino acids. Either of these diets will stimulate insulin secretion and thus insulin signaling.
Insulin signaling coordinately increases synthesis of uncharged tRNAs and ATF4. ATF4 increases the synthesis of nonessential amino acids and the capacity of the cell to take up both nonessential and essential amino acids. However, cellular . Following incubation, total cellular RNA was analyzed by qPCR, and transcript levels were normalized to the transcript levels in cells incubated in the absence of insulin and histidinol, which were set at 1. Asns, asparagine synthetase; Slc7a1, the solute carrier family 7, member 1 (also known as CAT-1); Slc7a5, the solute carrier family 7, member 5 (also known as LAT-1); Slc1a4, the Slc1a4, solute carrier family 1, member 4 (also known as ASCT1). C, mouse L cells were transfected with the indicated siRNA duplex and then switched to medium C containing dexamethasone (10 nM) for 48 h. Cells were then incubated for an additional 6 h in the absence or presence of 100 nM insulin (upper panel) or 100 nM insulin and 2 mM histidinol (2 mM). Total cellular RNA was analyzed by qPCR, and transcript levels were normalized to the transcript levels in cells transfected with control siRNA but incubated in the absence of insulin and histidinol, which were arbitrarily set at 1. Data represent the average values obtained from at least three independent experiments, and error bars represent S.E. LDL-R, the low density lipoprotein receptor; HMGCoAR, HMG-CoA reductase; PDK4, pyruvate dehydrogenase kinase, isoenzyme 4. uptake of amino acids also requires that amino acids are present in the extracellular environment.
If the diet contains a high level of amino acids, then the combination of insulin signaling, dietary amino acids, and insulinmediated ATF4 expression will increase intracellular amino acids. Under these conditions, uncharged tRNAs will tend to be converted into charged tRNAs, the GAAC pathway (which is activated by uncharged tRNAs) will tend to remain inactive, and high levels of charged tRNAs will promote anabolism.
On the other hand, if the diet contains a low level of amino acids, intracellular amino acid levels will fall, leading to a rise in uncharged tRNA levels and activation of the GAAC pathway (4). Co-activation of the insulin signaling and GAAC pathways would enhance the expression of ATF4 and downstream amino acid transporters and amino acid biosynthetic enzymes. We hypothesize that the combined additions of insulin and histidinol mimic this scenario.
If the combined effects of the GAAC and insulin signaling pathways are sufficient to increase intracellular amino acids, then uncharged tRNAs would be converted to charged tRNAs, the GAAC pathway would be inactivated, and anabolism would proceed. Alternatively, if the intracellular amino acid deficiency cannot be overcome, then persistent GAAC pathway activity would inhibit anabolism (4,29).
In summary, our data suggest that there are two biochemically distinct pathways to ATF4 expression, and we hypothesize that the GAAC pathway is active to varying degrees, depending on the amino acids that are available from the diet. The evolutionary conservation of ATF4 and the development of complex regulatory mechanisms to control ATF4 are consistent with the critical role of amino acids in cellular function. Our data suggest that in mammals, ATF4 is under dual control: by the ancient, cell autonomous GAAC pathway; and by signaling pathways that evolved later and brought key anabolic transcription fac-tors such as ATF4 under control of metabolic hormones such as glucocorticoids and insulin.