Amino Acid Deprivation Induces the Transcription Rate of the Human Asparagine Synthetase Gene through a Timed Program of Expression and Promoter Binding of Nutrient-responsive Basic Region/Leucine Zipper Transcription Factors as Well as Localized Histone Acetylation*

Expression of human asparagine synthetase (ASNS), which catalyzes asparagine and glutamate biosynthesis, is transcriptionally induced following amino acid deprivation. Previous overexpression and electrophoresis mobility shift analysis showed the involvement of the transcription factors ATF4, C/EBPβ, and ATF3-FL through the nutrient-sensing response element-1 (NSRE-1) within the ASNS promoter. Amino acid deprivation caused an elevated mRNA level for ATF4, C/EBPβ, and ATF3-FL, and the present study established that the nuclear protein content for ATF4 and ATF3-FL were increased during amino acid limitation, whereas C/EBPβ-LIP declined slightly. The total amount of C/EBPβ-LAP protein was unchanged, but changes in the distribution among multiple C/EBPβ-LAP forms were observed. Overexpression studies established that ATF4, ATF3-FL, and C/EBPβ-LAP could coordinately modulate the transcription from the human ASNS promoter. Chromatin immunoprecipitation demonstrated that amino acid deprivation increased ATF3-FL, ATF4, and C/EBPβ binding to the ASNS promoter and enhanced promoter association of RNA polymerase II, TATA-binding protein, and TFIIB of the general transcription machinery. A time course revealed a markedly different temporal order of interaction between these transcription factors and the ASNS promoter. During the initial 2 h, there was a 20-fold increase in ATF4 binding and a rapid increase in histone H3 and H4 acetylation, which closely paralleled the increased transcription rate of the ASNS gene, whereas the increase in ATF3-FL and C/EBPβ binding was considerably slower and more closely correlated with the decline in transcription rate between 2 and 6 h. The data suggest that ATF3-FL and C/EBPβ act as transcriptional suppressors for the ASNS gene to counterbalance the transcription rate activated by ATF4 following amino acid deprivation.

Mammalian cells have evolved complex cellular responses to changes in environment, including nutrient availability. The asparagine synthetase gene, ASNS, encoding the enzyme that catalyzes the synthesis of asparagine and glutamate using glutamine and aspartate, is a gene for which transcription is highly regulated by the nutritional status of the cell (1). Promoter analysis of the human ASNS gene by Barbosa-Tessmann et al. (2) demonstrated the presence of two cis-elements, termed nutrient-sensing response elements (NSRE-1, nt Ϫ68 to Ϫ60; NSRE-2, nt Ϫ48 to Ϫ43) in the ASNS promoter region. Both of these elements are essential for transcriptional activation through either the amino acid response (AAR) 1 pathway following amino acid deprivation or the endoplasmic reticulum stress response pathway, also known as the unfolded protein response, following glucose deprivation or other ER stress conditions. Although the NSRE-2-binding proteins are unknown, transient expression and electrophoresis shift analysis have demonstrated that activation of the ASNS gene involves activating transcription factor 4 (ATF4) (3), ATF3 (4), and CCAAT/ enhancer-binding protein ␤ (C/EBP␤) (5) action via the NSRE-1 site.
ATF and C/EBP proteins are subfamilies of the larger bZIP (basic region/leucine zipper) transcription factor family. ATF4 is widely expressed in a variety of tissues and tumor cell lines (6). Both transcription (3) and translation (7,8) of ATF4 are selectively increased in stress conditions, even when global protein synthesis is repressed, resulting in the induction of ATF4 target genes such as ASNS, CHOP, and ATF3. Translation from pre-existing ATF4 mRNA is enhanced in stress conditions that lead to eIF-2␣ phosphorylation, including amino acid deprivation (9), ER stress (10), the presence of long double strand RNA (11), and heme deficiency (12). ATF3 is expressed at low levels in normal and quiescent cells but can be rapidly induced in response to diverse stress signals and is likely to be involved in controlling a wide variety of cellular activities (13,14). As the result of alternative splicing, activated transcription from the ATF3 gene results in several different mRNA 1 The abbreviations used are: AAR, amino acid response; AARE, amino acid response element; ATF, activating transcription factor; C/EBP, CCAAT/enhancer-binding protein; NSRE, nutrient sensing response element; RT, reverse transcription; qPCR, quantitative realtime PCR; qRT-PCR, quantitative real-time RT-PCR; ChIP, chromatin immunoprecipitation; ER, endoplasmic reticulum; ERSE, ER stress response element; CMV, cytomegalovirus; CHOP, C/EBP homology protein; MEM, minimal essential medium; nt, nucleotide(s); pol II, polymerase II; HAT, histone acetyltransferase; PCAF, p 300 /CREB-associated factor. species, encoding proteins of different sequence, coding frames, and length (4, 14 -16). The longest protein, ATF3-full length (ATF3-FL) can function as a homodimer, in which case it often acts to repress transcription, or as a heterodimer with other bZIP family members, in which case it can either repress or activate transcription (6). Pan et al. (4) and Jiang et al. (17) demonstrated that the expression of ATF3 is induced in response to amino acid deprivation or to ER stress, by mechanisms requiring the eIF-2␣ kinases GCN2 and PERK, respectively. Pan et al. (4) also documented that histidine deprivation leads to alternative splicing of ATF3 such that an mRNA encoding a truncated ATF3 protein is favored. C/EBP␤ is a member of a transcription factor family, including C/EBP␣, ␤, ␥, ␦, ⑀, and CHOP. C/EBP␤ is involved in a wide range of cellular processes, such as cellular proliferation, carbohydrate metabolism, adipocyte differentiation, and inflammation (18). Marten et al. (19,20) reported that the mRNA content of C/EBP␤ is increased by amino acid deprivation. Subsequently, Siu et al. (5) showed that not only is C/EBP␤ mRNA increased by histidine deprivation of HepG2 hepatoma cells, but there is also increased DNA binding activity for C/EBP␤ in the nuclear extract.
The present study was designed to further investigate the expression levels of ATF4, ATF3, and C/EBP␤ protein following amino acid limitation and to establish the temporal interactions of ATF4, ATF3, and C/EBP␤ with the ASNS promoter using chromatin immunoprecipitation (ChIP) analysis. Furthermore, we have observed that a change of histone acetylation status occurs within the ASNS promoter region when the gene is activated following amino acid deprivation. Reversible acetylation of nucleosomal core histones is correlated with changes in chromatin during transcription. Specifically, actively transcribed genes are commonly featured by hyperacetylation of histones H3 and/or H4 tails located within their promoter regions (21). Using ChIP analysis, a time course of following amino acid limitation revealed a rapid increase, within 45 min, in ATF4 binding and acetylation of histones H3 and H4. Together with the association of the general transcription initiation complex, these changes closely paralleled the increase in transcription rate from the ASNS gene. Subsequently, there was increased binding of ATF3-FL and C/EBP␤ to the ASNS promoter, which occurred between 2 and 4 h and was more closely associated with the decline in transcription rate. The results demonstrate that following amino acid limitation there is a sequential order of association for a number of important bZIP transcription factors that bind to the ASNS promoter NSRE-1 site and that this factor binding is accompanied by modification of chromatin.
Cell Culture-Human HepG2 hepatoma cells were cultured in minimal essential medium (MEM, pH 7.4, Mediatech, Herndon, VA), supplemented to contain 1ϫ non-essential amino acids, 4 mM glutamine, 100 g/ml streptomycin sulfate, 100 units/ml penicillin G, 0.25 g/ml amphotericin B, and 10% (v/v) fetal bovine serum (FBS). Cells were maintained at 37°C in a 5% CO 2 /95% air incubator. Given that ASNS promoter activity is induced by nutrient depletion, cell cultures were replenished with fresh MEM medium and serum for 12 h prior to initiating all treatments to ensure that the cells were in the basal ("fed") state. Amino acid deprivation was performed by incubating the cells for 12 h in complete MEM or MEM lacking histidine, each containing 10% dialyzed fetal bovine serum.
RNA Isolation and Real-time Quantitative RT-PCR-Total cellular RNA was isolated from HepG2 cells using the Qiagen RNeasy kit (Qiagen). For each sample, 2 g of RNA was dried and resuspended in RNase-free H 2 O to make a final concentration of 200 ng/l. To measure the relative amount of ASNS mRNA, quantitative real-time RT-PCR (qRT-PCR) analysis was performed using a DNA Engine Opticon 2 system (MJ Research, Reno, NV) and detection with SYBR Green I. The primers for amplification were: sense primer, 5Ј-GCAGCTGAAA-GAAGCCCAAGT-3Ј and antisense primer, 5Ј-TGTCTTCCATGCCAAT-TGCA-3Ј. The reactions were incubated at 50°C for 30 min followed by 95°C for 15 min to activate the Taq polymerase and amplification of 35 cycles of 95°C for 15 s, and 60°C for 60 s. After PCR, melting curves were acquired by stepwise increase of the temperature from 55°C to 95°C to ensure that a single product was amplified in the reaction. L7a mRNA level was also measured at the same time as the internal control. The primers for amplification were: sense primer, 5Ј-TTTGGCATTG-GACAGGACATCC-3Ј; and antisense primer, 5Ј-AGCGGGGCCATT-TCACAAAG-3Ј. PCR was done in duplicates with samples from at least three independent experiments, and the means Ϯ the standard error of the means between conditions were compared by Students t test.
Transcription Rate Determination-Total RNA was isolated from HepG2 cells using the Qiagen RNeasy kit (Qiagen), including DNase I treatment before final elution to eliminate any DNA contamination. To measure the transcription rate from the ASNS gene, oligonucleotides derived from ASNS intron 12 and exon 13 were used to measure unspliced transcript (pre-mRNA). This procedure for measuring transcription rate is based on that described by Lipson and Baserga (22), except that pre-mRNA levels were analyzed by qRT-PCR. Reactions without reverse transcriptase were performed as a negative control to rule out any amplification from any residual genomic DNA. These tests were always negative. The primers for amplification were: sense primer, 5Ј-CCTGCCATTTTAAGCCATTTTGC-3Ј; and antisense primer, 5Ј-TGGGCTGCATTTGCCATCATT-3Ј. The reactions were incubated at 50°C for 30 min followed by 95°C for 15 min to activate the Taq polymerase and amplification of 35 cycles of 95°C for 15 s, and 58°C for 60 s. After PCR, melting curves were acquired by stepwise increase of the temperature from 55 to 95°C to ensure that a single product was amplified in the reaction.
Immunoblotting-Total cell extracts or nuclear extracts were prepared at the time points indicated, and 30 g/sample was separated on a 4 -20% Tris-HCl polyacrylamide gel (Bio-Rad) or a standard 20-cm 10% gel (for C/EBP␤) and then electrotransferred to a Protran nitrocellulose membrane (Schleicher & Schuell). The membrane was stained with Fast Green to check for equal loading and then incubated with 10% blocking solution (10% (w/v) Carnation nonfat dry milk, 30 mM Tris-Base, pH 7.5, 0.1% (v/v) Tween 20, and 200 mM NaCl) for 2 h at room temperature with mixing. Immunoblotting was performed using rabbit polyclonal antibodies against ATF3-FL, ATF4, or C/EBP␤ (all isoforms) at an antibody concentration of 0.2-0.4 g/ml in 10% dry milk blocking solution for 2 h at room temperature (ϳ21°C). The blots were washed 5 ϫ 5 min in 5% blocking solution on a shaker and then incubated with peroxidase-conjugated goat anti-rabbit secondary antibody (Kirkegaard & Perry Laboratories, Gaithersburg, MD) at a 1:20,000 dilution for 45 min at room temperature. The blots were then washed for 5 ϫ 5 min in 5% dry milk blocking solution and 2 ϫ 5 min in freshly made TBS/ Tween (30 mM Tris-base, 0.1% Tween 20, and 200 mM NaCl, pH 7.5). The bound secondary antibody was detected using an enhanced chemiluminescence kit (Amersham Biosciences) and exposing the blot to BioMax MR film (Kodak, Rochester, NY).
Transient Transfection and Expression-The reporter plasmid (ASNS Ϫ173/ϩ51 /luc) was made by inserting the ASNS promoter region (nt Ϫ173/ϩ51) upstream of the Firefly luciferase reporter gene using the HindIII site of the pGL3 plasmid (Promega, Madison WI). The expression of the rat ATF4 cDNA (provided by Dr. Jawed Alam, Alton Ochsner Medical Foundation, New Orleans, LA), ATF3-FL cDNA (provided by Dr. Tsonwin Hai at The Ohio State University), and C/EBP␤-LAP cDNA (provided by Dr. U. Schibler via Dr. Harry S. Nick, University of Florida) was driven by the cytomegalovirus promoter. HepG2 cells (0.2 ϫ 10 6 cells/well) were seeded on 24-well plates 18 -24 h before transfection with Superfect reagent (Qiagen) at a ratio of 6 l of Superfect to 1 g of DNA. For each transfection, 0.5 g of the ASNS Ϫ173/ ϩ51/luc plasmid was used along with the indicated amounts of the transcription factor expression plasmids. The total amount of transfected DNA was kept constant among experimental groups by the addition of empty pcDNA3.1 plasmid. Eighteen hours following transfection, the cells were transferred to fresh complete MEM for 12 h before preparing cellular extracts for analysis of luciferase activity. To prepare the extract, cells were washed with phosphate-buffered saline and treated with 100 l of Passive Lysis buffer supplied by Promega. The lysates were collected and stored at Ϫ80°C until use. Twenty microliters of cell extract was used for Firefly luciferase assays using the Luciferase Reporter Assay System (Promega), and the luminescence was measured with a luminometer. The luciferase values are normalized to microgram protein content for each sample. For each experimental condition, four to six assays were performed for each transfection and at least three independent transfections were done.
ChIP Analysis-ChIP analysis was performed according to a modified protocol of Upstate Biotechnology, Inc. (Charlottesville, VA). Briefly, HepG2 cells were seeded at 1.5 ϫ 10 7 /150-mm dish with complete MEM and grown for 24 h. Cells were transferred to fresh MEM 12 h before transfer to either complete MEM or MEM lacking histidine (MEM-His) for the time period indicated in each figure. Protein-DNA was cross-linked by adding formaldehyde directly to the culture medium to a final concentration of 1% and then stopped 10 min later by adding 2 M glycine to a final concentration of 0.125 M. Cross-linked chromatin was solubilized by sonication using a Sonic Dismembrator (Model 60, Fisher Scientific Co.) for five bursts of 40 s at power 10 with 2-min cooling on ice between each burst. Extract from 1 ϫ 10 7 HepG2 cells was incubated with 2 g of antibody. A rabbit anti-chicken IgG was used as the nonspecific antibody control. The antibody-bound complex was precipitated by protein A-Sepharose beads (Amersham Biosciences). The DNA fragments in the immunoprecipitated complex were released by reversing the cross-linking at 65°C for 5 h and purified using a QIAquick PCR purification kit (Qiagen). Purified, immunoprecipitated DNA was first analyzed by PCR and visualized by ethidium bromide staining after gel electrophoresis. Primers used for the regular PCR were: forward primer (human ASNS Ϫ173), 5Ј-CAAAAGAGCTC-CTCCTTG; and reverse primer (human ASNS ϩ51), 5Ј-TAAGCAGGT-CAGGGTGAT. The results were confirmed by quantitative real-time PCR (qPCR) with primers (5Ј-TGGTTGGTCCTCGCAGGCAT-3Ј and 5Ј-CGCTTATACCGACCTGGCTCCT-3Ј) designed using Vector NTI version 7.1 software (InforMax Inc., Frederick, MD) to amplify the ASNS proximal promoter sequence nt Ϫ87 to Ϫ22 (5Ј-TGGTTGGTCC-TCGCAGGCATGATGAAACTTCCCGCACGCGTTACAGGAGCCAGG-TCGGTATAAGCG-3Ј, NSRE-1 and NSRE-2 are underlined). The qPCR analysis was performed using the DNA Engine Opticon 2 system and detected with SYBR Green I. Serial dilutions of input chromatin were used to generate a standard curve for determining the relative amount of product. Duplicates for both the standards and the samples were simultaneously amplified using the same reaction master mixture. The reactions were incubated at 95°C for 15 min to activate the polymerase, followed by amplification at 95°C for 15 s and 61.4°C for 60 s for 35 To test the effect of histidine limitation on the turnover of ASNS mRNA, HepG2 cells were incubated in MEM lacking histidine for 12 h and then transferred to fresh complete MEM or MEM lacking histidine, both containing 5 M actinomycin D (ActD) (B). At the indicated times, RNA was isolated and subjected to quantitative RT-PCR analysis for ASNS and L7a mRNA. The ASNS/L7a ratio is expressed as a percentage of the value obtained for time ϭ 0 (12 h of histidine starvation), and then plotted as the semi-log to reflect the decay rate of ASNS mRNA. To investigate transcription rate, HepG2 cells were incubated in histidine-free MEM for 0 -24 h for the long time course (C) and for 0 -2 h for the short time course (D). The transcription rate was determined by quantitative RT-PCR analysis of ASNS pre-mRNA with the specific primers spanning an intron-exon junction (as described under "Materials and Methods") to amplify transient intermediates, which reflect the transient unspliced transcripts. The graphs in C and D are the summary of three independent experiments, from which the data are depicted as the means Ϯ S.E. of the means.
cycles. After PCR, melting curves were acquired by stepwise increases in the temperature from 55 to 95°C to ensure that a single product was amplified in the reaction. The results are expressed as the ratio to input DNA. Samples from at least three independent immunoprecipitations were analyzed, and the means Ϯ S.E. of the means between conditions were compared by the Students t test.

Effect of Amino Acid Deprivation on ASNS mRNA Synthesis
and Turnover-Although it is known that the ASNS mRNA content of cells is increased after 12 h of amino acid limitation, a longer period of study is necessary to determine if this increase is transient. Consequently, HepG2 human hepatoma cells were incubated in histidine-free MEM for 0 -24 h and the mRNA level for ASNS was measured for three independent experiments by qRT-PCR (Fig. 1A). An initial increase in ASNS mRNA content was observed after 2 h of histidine deprivation and then reached a maximum of about 15-times the control value at 12 h. Although there was a trend of decline from 12 to 24 h, the level was still elevated by 10-times the control.
Although it has been documented that increased transcription contributes to the increase in ASNS mRNA (2), Gong et al. (23) reported that amino acid limitation also caused the increased stability of ASNS mRNA in Chinese hamster ovary cells. To evaluate whether mRNA stabilization contributes to the increase in ASNS mRNA in HepG2 cells, they were incubated in MEM without histidine for 12 h to elevate ASNS mRNA content and then transferred to either fresh histidinefree MEM or complete MEM, both containing 5 M actinomycin D (Fig. 1B). ASNS RNA content was assayed by qRT-PCR, and the rate of decay was analyzed graphically. It was clear that the rate of mRNA turnover was not different in the presence or absence of histidine, and in both conditions, the estimated mRNA half-life was ϳ20 h. These results indicate that amino acid deprivation of HepG2 cells does not elevate the ASNS mRNA content by increasing mRNA stability.
To investigate the change in the transcription rate of the ASNS gene, the levels of unspliced ASNS pre-mRNA were Immunoblots were probed with antibodies against ATF4, ATF3-FL, or C/EBP␤. To show equal lane loading and to control for nonspecific variation, blots were stained with Fast Green (not shown) and probed for ATF2 as a negative control (A). Overexpression of a cDNA encoding C/EBP␤-LAP and -LIP was used to identify the migration of these two isoforms (not shown). C/EBP␤-LAP migrated as three primary bands, as described under "Results," so each band was quantified separately. A and C shows representative autoradiograms, whereas the graphs in B and D are the summary of three independent experiments, from which the data are depicted as the averages Ϯ S.D. measured. Given that introns are rapidly removed from hnRNA during splicing, this procedure is a measure of transcription, similar to the data obtained by nuclear run-on analysis (22). HepG2 cells were incubated in histidine-free MEM for 0 -24 h for a long time course and for 0 -2 h for a shorter time course. qRT-PCR analysis of ASNS pre-mRNA with specific primers spanning an intron-exon junction was used to amplify a transient intermediate. In the long time course (Fig. 1C), an increase in ASNS transcription rate was observed after 2 h of histidine deprivation, which reached a maximum of about 20times the control value. There was a gradual decline from 2 to 24 h, although even after 24 h the level was still elevated compared with the control. When the first 2 h were examined (Fig. 1D), it was determined that the ASNS transcription rate started to increase after 45 min of histidine limitation and escalated to a maximum of 20-times the control value at 2 h.
ATF3-FL, ATF4, and C/EBP␤ Protein Content Responds to Amino Acid Limitation-To establish the kinetics of ATF4, ATF3-FL, or C/EBP␤ protein content after amino acid limitation, whole cell extracts from control (MEM-incubated) or histidine-deprived HepG2 cells were subjected to immunoblotting (Fig. 2). Amino acid deprivation caused an elevation of both ATF4 and ATF3-FL proteins (Fig. 2, A and B). Consistent with the recognized translational control of ATF4 by amino acid availability (7,8), ATF4 protein content was significantly increased by 2 h (Fig. 2, A and B), peaked at 4 h, and then declined gradually over the next 20 h to the control level. In contrast, the peak of ATF3-FL protein expression was later and occurred at 8 h, but stayed relatively high for the remainder of the 24-h period investigated (Fig. 2, A and B). The changes in C/EBP␤ protein composition were complex. The C/EBP␤ mRNA encodes three possible proteins, LAP*, LAP, and LIP depending on translational initiation at one of three methionine codons that exist within the sequence. LAP* and LAP are two different forms of "liver activating protein," whereas LIP, "liver inhibitory protein," is a truncated isoform that serves as a naturally occurring dominant negative (24). The predicted molecular masses are 36, 33.5, and 16.5 kDa for the human proteins, respectively, although they are known to run anomalously on SDS-PAGE and are subject to post-translational modification (18,25). To provide a standard reference, a human cDNA beginning with the second methionine codon was expressed in HepG2 cells, and thus, should yield only the LAP and LIP isoforms. Upon immunoblotting with an antibody that will recognize all C/EBP␤ isoforms, the LAP protein was detected as a series of at least three bands at ϳ40 -45 kDa (data Immunoblots were probed with antibodies against ATF4, ATF3-FL, or C/EBP␤. To test for equal lane loading and nonspecific variation, the blots were stained with Fast Green (not shown) and probed for ATF2 as a negative control (A). A and C shows representative autoradiograms, whereas the graphs in B and D are the summary of three independent experiments, from which the data are depicted as the averages Ϯ S.D. The dotted line represents the rate of transcription from the ASNS gene after histidine deprivation, and is taken from Fig. 1C. not shown). An immunoreactive band corresponding to LIP (ϳ17 kDa) was observed only when the blot was developed for a longer period, consistent with a reported lower rate of initiation at the last methionine codon (26). The endogenous C/EBP␤-LAP migrated as a series of bands that corresponded to those obtained with the overexpressed standard (Fig. 2C,  bands A-C). The most intensive band detected in the control cells (MEM) actually decreased slightly during the course of histidine deprivation (Fig. 2C, LAP-B band). However, the band above (A Band) and the band below (C Band) the band B were increased in abundance following amino acid limitation. Although the absolute amount of these protein forms was less than band B, the relative increase was significant (Fig. 2D). The molecular basis for this heterogeneity remains to be established, but C/EBP␤ is known to be the target of post-translational modification (18,25).
To determine whether or not ATF4, ATF3-FL, or C/EBP␤ protein content in the nucleus was altered, immunoblots of nuclear extracts from control or histidine-deprived HepG2 cells were also subjected to immunoblotting (Fig. 3). Amino acid deprivation caused an elevation of both ATF4 and ATF3-FL proteins in the nucleus. The increase in nuclear ATF4 content (Fig. 3, A and B) occurred earlier than that in the whole cell extract, suggesting that the initial increase in ATF4 protein is rapidly transferred to the nuclear compartment. Consistent with the hypothesis that ATF4 is a key regulator of amino acid-dependent transcription, the increase in nuclear ATF4 content closely paralleled the increased transcription rate of the ASNS gene (Fig. 3B, dotted line). In contrast, the change in the ATF3-FL protein level (Fig. 3, A and B) was similar to that in the whole cell extract, and the ATF3 nuclear content correlated with the decline in transcription rate that occurred between 4 and 12 h. The initial changes in nuclear C/EBP␤ protein (Fig. 3, C and D) paralleled those in the whole cell extract, but after 12 h the nuclear increase in bands A and C was more transient when compared with that in the whole cell extract. As with ATF3, the greatest increase in C/EBP␤ bands A and C occurred at a time (4 -12 h) when the rate of transcription (dotted line in Fig. 3D) (4) individually resulted in regulated transcription from a reporter gene driven by the ASNS promoter. However, the observed temporal differences for expression of ATF4, C/EBP␤, and ATF3-FL proteins suggest that their effect on the ASNS promoter activity is likely complex. The coordinated control of ASNS by ATF4, ATF3-FL, and C/EBP␤ was investigated by transfecting different combinations of the expression vectors for ATF4, C/EBP␤, or ATF3-FL and the influence of protein content was assessed by transfecting different concentrations of plasmid DNA (Fig. 4). When HepG2 cells were transfected with ATF4 alone, at concentrations of 10 or 100 ng of plasmid DNA, ASNS-driven transcription was enhanced in a concentration-dependent manner, with a maximum induction of about 30-times the MEM control value at 100 ng of ATF4. When ATF4 was co-transfected along with the ATF3-FL and C/EBP␤, the latter two factors resulted in an inhibition of the ATF4induced transcription. At 10 ng of ATF4 cDNA, increasing the C/EBP␤ amount from 10 to 100 ng, but leaving ATF3-FL at 10 ng, resulted in about 50% suppression (Fig. 4). However, when the ATF3-FL amount was raised to 100 ng, nearly a complete inhibition of ASNS promoter activity was observed (Fig. 4). However, at the higher ATF4 plasmid concentration (100 ng), ATF4-induced ASNS promoter activity was suppressed by the combination of ATF3-FL and C/EBP␤, but to a lesser extent. The amount of ATF3-FL appeared to be more important than C/EBP␤, in that 100 ng of ATF3-FL largely blocked the low concentration ATF4-induced transcription, regardless of the amount of C/EBP␤. Consistent with this interpretation, when 10 or 100 ng of C/EBP␤ plasmid was tested in the absence of ATF3, a modest enhancement rather than inhibition of ATF4 action was observed (data not shown). Collectively, the results show that when present together ATF3-FL and C/EBP␤ antagonize the ATF4-mediated activation of the ASNS promoter.

Identification of Transcription Factors Interacting in Vivo with the ASNS Promoter Region during Amino Acid Deprivation and Recruitment of RNA Polymerase II-Electrophoresis
mobility shift analysis has demonstrated that C/EBP␤, ATF4, and ATF3-FL are each capable of binding to the NSRE-1 sequence within the ASNS promoter (3)(4)(5). Conversely, several other members of the ATF and C/EBP families bind weakly, or not at all. To test ATF and C/EBP family member interaction with the ASNS promoter in vivo during amino acid deprivation,

FIG. 4. ATF4, ATF3-FL, and C/EBP␤ coordinately regulate ASNS promoter activity in vitro.
To investigate the possible coordinated control of ASNS transcription by ATF4, ATF3-FL, and C/EBP␤, HepG2 cells were co-transfected with an ASNS Ϫ173/ϩ51 promoter/luc reporter plasmid, along with expression vectors for ATF4 alone, or ATF4 in combination with ATF3-FL or C/EBP␤ at different plasmid concentrations. The total amount of transfected DNA was kept constant among experimental groups by the addition of pcDNA3.1 plasmid DNA. Cell extracts were assayed for luciferase activity, as detailed under "Materials and Methods." Each experiment was repeated with multiple batches of cells, and the data are presented as the means Ϯ S.E. of the means. chromatin immunoprecipitation (ChIP) assays were performed using antibodies specific for ATF1, ATF2, ATF3-FL, ATF4, C/EBP␣, C/EBP␤, C/EBP␦, and C/EBP⑀. A rabbit polyclonal antibody against chicken IgG was used as nonspecific control (n/s IgG). After isolation of cross-linked chromatin from cells incubated in either complete medium (MEM) or MEM lacking histidine (MEM-His), immunoprecipitated DNA was analyzed by both regular PCR (ethidium bromide-stained gels) and qPCR. Gel electrophoresis of PCR products showed a single product as expected and showed that after an 8-h incubation in MEM-His there was an apparent increase in ATF3-FL, ATF4, and C/EBP␤ binding to the promoter (Fig. 5A). Increased binding of RNA polymerase II (pol II) to the ASNS promoter during amino acid deprivation was also confirmed (Fig. 5A). The in-

FIG. 5. Association of transcription-related proteins with the promoter region of ASNS gene after amino acid deprivation.
Chromatin immunoprecipitation assays were performed as described under "Materials and Methods." To collect the data of A, HepG2 cells were treated with either complete MEM (M) or MEM lacking histidine (ϪH) for 8 h before cross-linking and immunoprecipitation by antibodies against the indicated proteins. PCR products amplifying nt Ϫ173 to ϩ51 of the human ASNS gene were size-fractioned by gel electrophoresis and representative gels are shown (A). A rabbit anti-chicken IgG was used as the nonspecific negative control (n/s IgG). Dilutions of input DNA were tested as template for PCR to show linear amplification. For the data in B-D, quantitative real-time PCR was performed with immunoprecipitated DNA and a 1:20 dilution of input DNA samples using primers to amplify nt Ϫ87 to Ϫ22 of the human ASNS gene. The amino acid-responsive NSRE-1 regulatory site is located at nt Ϫ68/Ϫ62 (sequence given under "Materials and Methods"). Data were plotted as the ratio to the value obtained with the 1:20 dilution of input DNA. The single asterisks indicate those MEM-His samples that had statistical significance with a p Յ 0.05 compared with the control (MEM, B and C). For D, none of the MEM-His samples were statistically different from the MEM controls, but there was constitutive binding and the double asterisks indicate statistical significance with p Յ 0.05 compared with the correspondent nonspecific IgG samples. crease in ASNS promoter binding by ATF and C/EBP family members, pol II and general transcription factors, and acetylated histones in amino acid-deprived cells was quantified by real-time qPCR analysis (Fig. 5, B-D). Among the other members of the ATF and C/EBP families tested, binding of ATF1 and C/EBP⑀ only occurred at a level approximately equal to the nonspecific IgG (data not shown). On the other hands, ATF2, C/EBP␣, and C/EBP␦ promoter binding was significantly higher than the background value (nonspecific IgG), and for ATF2 there was no difference in the level of binding between control and histidine-deprived cells (Fig. 5D). For C/EBP␣ and C/EBP␦ the amount of binding in the amino acid-deprived cells tended to be elevated (by 2-fold), but the difference did not reach statistical significance (Fig. 5D). Although not responsible for mediating the amino acid-dependent transcription, three GC-rich sequences, just upstream from the NSRE-1 site, do function to maintain the basal rate of ASNS transcription and act in a permissive manner to allow a maximum response to the AAR and unfolded protein response pathways (27). Transient expression studies in Sp-deficient SL2 cells demonstrated that the maintenance of basal transcription could be supported by either Sp1 or Sp3. However, of these two factors, only Sp3 would permit the increase in transcription following amino acid limitation. Although detectable, the amount of Sp1 binding to the ASNS promoter was not statistically different from the nonspecific IgG (Fig. 5D). Consistent with the transient expression data, there was a significant amount of Sp3 binding in HepG2 cells incubated in either MEM or MEM-His (Fig. 5D).
Association of the Transcription Initiation Complex at the ASNS Promoter During Amino Acid Deprivation-To investigate the temporal association of pol II and the general transcription machinery with the ASNS promoter, HepG2 cells were incubated in either MEM or MEM-His for 0 -8 h, and ChIP analysis was performed with cells collected from specific time periods. Data were normalized against the corresponding input DNA data to eliminate any differences in cell numbers between samples. The results show that an increased association of pol II with the ASNS promoter occurred within 1 h after transfer to the histidine-deprived medium and was maintained at a high level throughout the remainder of the 8-h amino acid deprivation period (Fig. 6). The cells incubated in complete MEM medium showed a minimal level of pol II association without any change throughout the time course. When the transcription rate was overlaid on the same graph (Fig. 6,  dotted line), the results showed that the association of RNA pol II with the ASNS promoter closely paralleled the increased transcription rate of ASNS gene. In agreement with the binding pattern for RNA pol II, the interaction of TFIIB and TATAbinding protein, components of the general transcription ma- chinery, increased in the amino acid-deprived cells during the early stage of the treatment and was maintained at a high level throughout. In contrast, analysis of the binding for another general transcription factor, TAFII250, showed no significant association (compared with the nonspecific IgG, Fig. 7) with the ASNS promoter in either MEM or MEM-His incubated cells (Fig. 6).

Binding of Selected Transcription Factors to the ASNS Promoter Is Associated with Acetylation of Local Histones During
Amino Acid Deprivation-Histone acetylation is associated with chromatin remodeling and increased transcription (28). To test for the influence of amino acid deprivation on chromatin remodeling, acetylation of histones within the ASNS promoter region was also analyzed. In an initial experiment, the abundance of acetylated histones H3 and H4 was significantly increased in amino acid-deprived cells when compared with control (MEM) cells at the single time point of 8 h (Fig. 5C). A more complete time course showed a significant increase in acetylated histone H3 and H4 at the promoter region within 1 h of removal of histidine from the medium (Fig. 6). The histones remained highly acetylated for the next several hours and then at 4 -8 h of amino acid deprivation treatment, the degree of acetylation dropped slightly, a result that correlated to the changes observed for the transcription rate (dotted line in the AcH3 panel). On the other hand, qPCR using primers that amplify a region of exon 7 within the ASNS gene showed little or no histone acetylation at any of the times studied (data not shown). Therefore, the amino acid deprivation-induced histone acetylation appeared to be specific to the promoter region of the ASNS gene and correlated with activation of the gene. Interestingly, ChIP showed no significant recruitment to the ASNS promoter of proteins with known histone acetyltransferase activity, including CBP, GCN5, p300, or PCAF (data not shown). Consequently, the identity of the histone acetyltransferase that acts on the ASNS gene following amino acid limitation is an interesting aspect of amino acid control that remains to be determined.
Binding of Selected Transcription Factors to the ASNS Promoter during Amino Acid Deprivation-To study the temporal association of ATF3-FL, ATF4, and C/EBP␤ with the ASNS promoter, the time course of factor binding was analyzed after transferring HepG2 cells to either fresh MEM or MEM-His medium. The nonspecific IgG antibody gave minimal signals through the course of treatments and showed no difference between MEM-and MEM-His-treated cells (Fig. 7). On the other hand, immunoprecipitation of the ASNS promoter by antibodies against ATF3-FL, ATF4, or C/EBP␤ showed a significantly elevated level of binding during amino acid deprivation compared with the complete MEM condition (Fig. 7). However, the transcription factors exhibited a distinct pattern of promoter binding. ATF4 rapidly reached a maximal level of binding within 2 h of amino acid removal and that high level The presence of NSRE-1-bound ATF4 (and likely other newly bound proteins) leads to localized histone acetylation, presumably through the recruitment of an unidentified histone acetyltransferase, which, in turn, leads to binding of the general transcription machinery and the transcription initiation complex, including RNA polymerase II. As the period of amino acid deprivation continues during hours 4 -12 (Phase II), transcription from the ATF3 gene is enhanced (4) and the ATF3 mRNA is stabilized (Y-X. Pan and M. S. Kilberg, unpublished observations), resulting in increased nuclear ATF3 protein levels. Within the same time frame, there is also a change in the ratio of C/EBP␤-LAP isoforms within the nucleus. As a result, binding of both ATF3 and C/EBP␤ to the ASNS promoter is increased and concurrently, there is a relative decline in the rate of transcription. was maintained throughout the 9 h investigated. For both C/EPB␤ and ATF3-FL, no increase in binding to the ASNS promoter was observed for the first hour, and then binding rose steadily during the amino acid deprivation period reaching a maximum at 4 -6 h (Fig. 7). By plotting the transcription rate (from Fig. 1B) on the same graph, it is apparent that the binding of ATF4 to the promoter, like pol II (Fig. 6), paralleled the elevated transcription rate, whereas the association of ATF3-FL and C/EBP␤ appeared to correlate with the decline in transcription rate (Fig. 7).
To further define the role of ATF4 on ASNS activation, a shorter time course (0 -2 h) of factor binding was analyzed to establish the earliest time of association of ATF4, pol II, and acetylated histones H3 and H4 with the ASNS promoter. The results established that each of these proteins exhibited a significantly elevated level of binding only 45 min following amino acid deprivation (Fig. 8), indicating that the association of ATF4 with the ASNS promoter, RNA pol II recruitment, and histone modification occur almost simultaneously following amino acid deprivation. Furthermore, when the transcription rate was measured over this shorter time period, factor binding closely paralleled the increased transcription rate of the ASNS gene (Fig. 8, dotted line).

DISCUSSION
The data in the present study show several novel observations that illustrate the role of ATF4, ATF3, and C/EBP␤ in regulating human ASNS gene expression following amino acid deprivation. 1) The protein content for ATF4 and ATF3-FL were increased during amino acid limitation, whereas C/EBP␤-LIP declined slightly. The total amount of C/EBP␤-LAP protein was unchanged, but changes in the distribution among multiple C/EBP␤-LAP forms were observed. Consistent with the known translational control of ATF4 synthesis (7,8), the increase in ATF4 protein occurred faster than that for ATF3-FL or the changes in C/EBP␤. 2) Transient overexpression of ATF3-FL and C/EBP␤ antagonized the ATF4-mediated activation of the ASNS promoter. 3) Following amino acid deprivation, ATF4, ATF3, and C/EBP␤ physically associated with the ASNS promoter in a time-dependent manner, along with binding of RNA polymerase II and other general transcription factors. 4) ChIP analysis also demonstrated for the first time the temporal changes in the local chromatin environment in terms of histone modification following amino acid depletion. 5) Using ChIP, a time course analysis of the first 2 h following amino acid removal revealed a sequential binding of transcription factors, with rapidly increased ATF4 binding, acetylation of histone H3 and H4 tails, and increased association of the general transcription initiation complex closely paralleling the elevated transcription rate. Increased binding of ATF3-FL and C/EBP␤ to the ASNS promoter occurred later and correlated with a decline in transcription rate.
Collectively, published data and the present observations have led to a working model, presented in Fig. 9, for activation of the human ASNS gene in response to amino acid limitation. Characterization of mammalian amino acid response elements, including NSRE-1, is still in the early stages, so the purpose of the model is to not only illustrate what is known but what important issues remain to be discovered. The present results provide the first in vivo evidence for factor binding to an AARE and illustrate the key role that ATF4 plays in activating the ASNS gene. The model also proposes that there is a selflimiting program that results in a delayed activation of transcription factors such as ATF3 and C/EBP␤, which appear to suppress the activation of the ASNS gene and make it transitory. This transition to a less activated state may serve to redirect the metabolic resources within the cell.
ATF4, a member of the ATF/CREB transcription factor family, has been shown to interact with RPB3, a core subunit of RNA polymerase II (pol II) (29). ATF4 transactivation is enhanced by RPB3, whereas when used as a dominant negative, the region of RPB3 that contacts ATF4 markedly inhibits ATF4 transactivating activity. In the present study, the temporal relationship of ATF4 binding, the recruitment of RNA polymerase II, and the transcription rate changes suggest that ATF4 is the main factor that triggers increased ASNS transcription shortly after amino acid limitation. The peak of ATF4 protein expression in the nucleus occurred at 2-4 h and then declined gradually over the next 20 h to the control level. Using a procedure to assay the levels of unspliced ASNS hnRNA as a measure of transcription rate, transcription factor binding and gene activation were precisely correlated. Following histidine limitation, ATF4 binding was significantly increased and closely paralleled the elevation in transcription rate.
The increase in ATF3-FL protein occurred more than 4 h after histidine removal and stayed relatively high for the remainder of the 24-h period investigated. The transcription analysis coupled with the measurements of ATF3-FL and C/EBP␤ binding to the ASNS promoter suggests that these two factors antagonize the ATF4-mediated activation of the ASNS promoter and may serve to suppress the transcription rate previously induced by amino acid limitation. The full-length ATF3 protein contains a C-terminal basic DNA recognition region and leucine zipper dimerization domain (bZip) characteristic of the bZIP superfamily of transcription factors. ATF3-FL can homodimerize, but it can also heterodimerize with c-Jun, JunB, JunD, ATF2, and CHOP to facilitate DNA binding to an ATF/CRE or AP-1 consensus site (6). The transcriptional consequences are different depending on whether ATF3-FL binds as a homodimer, in which case it appears to act to repress transcription, or as a heterodimer with other bZIP family members, in which case it can either repress or activate transcription (13). Therefore, depending on the cellular context, dimerization partner, and target promoter, ATF3 action must be established for each gene. The transient expression studies of ATF3-FL and C/EBP␤ suggest that, although C/EBP␤ may have a repressive effect on ATF4 function, it is likely that ATF3-FL serves as the primary antagonist. This observation is consistent with our earlier observations showing the ability of ATF3-FL to counterbalance ATF4 action (4). Thus, the combination of ATF3-FL and C/EBP␤ appear to provide transcriptional restraint on the activated ASNS gene.
Interestingly, by performing electrophoresis mobility shift analysis following arsenite-induced stress, Fawcett et al. (30) reported a transient increase in ATF4 binding to the C/EBP-ATF composite site in the human CHOP promoter at 2 h, which was subsequently replaced at about 6 h by ATF3 as transcription from the gene declined back toward the basal expression rate. They also showed that ATF4 activates the CHOP gene through the C/EBP-ATF composite site and that ATF3 antagonizes that induction. The present data extends the observations of Fawcett et al. (30) by showing that this sequence of events can be observed in vivo by ChIP analysis. In addition to ASNS and CHOP, C/EBP-ATF composite sites that function as AAREs have been reported for the SNAT2 neutral amino acid transporter (31) and the cationic amino acid transporter Cat-1 (32). Given the high degree of sequence similarity between NSRE-1 and the other C/EBP-ATF composite sites, it is possible that transient activation of these sites by ATF4 and subsequent suppression by ATF3, as proposed in our working model (Fig. 9), may be a general mechanism for control of these genes in response to a wide variety of cellular stresses.
With regard to other contributory factors to ASNS regulation, the ChIP analysis indicated ATF2, C/EBP␣, C/EBP␦, and Sp3 may constitutively bind to the promoter in both fed and amino acid deprived conditions. Therefore, association of these factors may participate in the response to amino acid deprivation, but the role of each has yet to be clarified. ATF2 has been implicated in the amino acid-dependent regulation of the CHOP gene and has been shown to be capable of binding to the CHOP AARE sequence in vitro (33). Averous et al. (34) recently reported that the phosphorylation state of ATF2 is an important factor in its action following amino acid deprivation.
The present experiments also demonstrated a significant change in promoter histone acetylation during amino acid deprivation. It is the first report indicating that changes in amino acid status induce histone modification. The kinetics of histone H3 and H4 acetylation at the ASNS promoter were nearly identical to those for ATF4 and RNA pol II association during amino acid deprivation. On the other hand, no acetylation of either H3 or H4 is observed in the coding region of the ASNS gene, documenting the localization of this response. This is in agreement with the belief that hyperacetylated histone H3 is concentrated at the 5Ј region of transcription start sites of actively transcribed genes in human genome but is greatly decreased downstream of the transcription start site (35). The acetylation of histone tails is thought to facilitate transcription by altering accessibility of DNA to transcriptional activators or chromatin remodeling complexes (reviewed in Ref. 36). It has been shown in recent studies that several transcriptional coactivators have intrinsic histone acetyltransferase (HAT) activity (37). For example, Gcn4, a yeast transcriptional activator and the counterpart to mammalian ATF4, recruits Gcn5 HAT complexes to selective promoters to induce local acetylation of histone H3 and subsequent transcription activation (38). Recruitment of co-activators with HAT activity such as p300, CBP, and PCAF, by a variety of transcription factors, is known to occur, and this HAT activity then mediates local acetylation of the local histone tails. Data presented here clearly document a temporal correlation of ATF4 binding with acetylation of histone H3 and H4 at the ASNS promoter region following amino acid limitation. Therefore, it is quite possible that ATF4 acts as the recruiting factor for an unknown HAT to make the ASNS promoter more accessible to pol II and the general transcription machinery.
The model presented in Fig. 9 illustrates the need for further investigation to discover the identity of the proteins that link the AARE-binding proteins (i.e. co-activators, co-repressors, etc.) to the general transcription machinery and the initiation complex, during both Phase I, the most activated state, and the more restrained period of Phase II. Investigation is underway to identify the one or more cofactors that interact with ATF4. Insight into how the transcriptional machinery assembles at the ASNS promoter and modulates transcription will provide additional understanding of the molecular steps required for nutritional control via the amino acid response pathway.