SIRT1 Suppresses Activating Transcription Factor 4 (ATF4) Expression in Response to Proteasome Inhibition

Abstract

The synthetic machinery of ATF4 (activating transcription factor 4) is activated in response to various stress conditions involved in nutrient restriction, endoplasmic reticulum homeostasis, and oxidation. Stress-induced inhibition of proteasome activity triggers the unfolded protein response and endoplasmic reticulum stress, where ATF4 is crucial for consequent biological events. In the current study, we showed that the $NAD^+$-dependent deacetylase, SIRT1, suppresses ATF4 synthesis during proteasome inhibition. SIRT1 depletion via transfection of specific siRNA into HeLa cells resulted in a significant increase in ATF4 protein, which was observed specifically in the presence of the proteasome inhibitor MG132. Consistent with SIRT1 depletion data, transient transfection of cells with SIRT1-overexpressing plasmid induced a decrease in the ATF4 protein level in the presence of MG132. Interestingly, however, ATF4 mRNA was not affected by SIRT1, even in the presence of MG132, indicating that SIRT1-induced suppression of ATF4 synthesis occurs under post-transcriptional control. Accordingly, we propose that SIRT1 serves as a negative regulator of ATF4 protein synthesis at the post-transcriptional level, which is observed during stress conditions, such as proteasome inhibition.

Introduction

Unfolded and misfolded proteins that are inappropriately processed in the endoplasmic reticulum (ER) are eliminated via proteasomal degradation. Attenuation or inhibition of proteasomal degradation activity leads to the accumulation of abnormal proteins in the ER, thereby eliciting ER stress [6,17,19]. Activating transcription factor 4 (ATF4), a member of the ATF/CREB (activating transcription factor/cyclic AMP response element binding protein) family, mediates responses to unfolded protein and ER stress [6].

The majority of normal cells exhibit not only inefficient translation of ATF4 mRNA but also active post-translational degradation of ATF4 protein [1]. Therefore, the basal level of ATF4 protein in normal cells remains too low for detection, despite the ubiquitous expression of ATF4 mRNA [1]. With the development of stress conditions, such as deprivation of amino acids, glucose, or oxygen, cells respond by accumulating ATF4 protein, mainly due to changes in translation and post-translational degradation, and partly due to transcription [2,5,11]. Under these stress conditions, translational inhibition of ATF4 mRNA by its upstream open reading frame is bypassed [1,10] and degradation of protein by E3 ligase β-TRCP reduced [1], leading to accumulation of ATF4 protein. ATF4 translation is basically regulated via phosphorylation of eIF2α [10]. Phosphorylation of eIF2α, induced during stress, preferentially increases ATF4 translation while suppressing that of most other mRNA m olecules [11,16,24,28]. Based o n the t ype of stress, different kinases are activated to trigger eIF2α phosphorylation, including PERK for endoplasmic reticulum stress, GCN2 for amino acid deprivation [26], PKR for viral infection [8], and HRI for heme deficiency [18]. Posttranslational modification of ATF4 contributes to protein degradation via modulation of β-TRCP interactions with the “DSGXXXS” motif on ATF4 [15]. As discussed above, the regulatory factors and mechanisms contributing to cellular ATF4 levels are diverse, depending on the type of stress and regulation steps. The regulatory factors underlying the complex physiologies associated with ATF4 remain to be established.

In the last decade, SIRT1, an NAD+-dependent deacetylase, has been identified as a sensor for detecting internal and external stress and delivering information to target proteins [20]. Accumulating evidence has shown that transcription factors are the main downstream targets for SIRT1. In addition, SIRT1 controls the activity of transcription factors by regulating the nuclear translocation of metabolic enzymes such as GAPDH [13,25]

The regulation of transcription factors by SIRT1 is achieved through the modulation of chromatin structure via direct interactions with histone leading to deacetylation [21]. Moreover, p300 acetyltransferase participates in these cascade pathways counteracting SIRT1. As indicated above, ATF4 is regulated at the transcriptional, translational, and post-translational levels. Identification of the controller molecules involved in each step is a key challenge. In addition, it is important to understand the precise mechanisms by which proteasome inhibition regulates ER stress [23]. Based on the finding that both SIRT1 and ATF4 control the same physiological processes (such as nutrient restriction and proteasome degradation) and interact with p300, we hypothesized that SIRT1 functions as a controller of ATF4 expression. Data from the present study showed that SIRT1 acts as a negative regulator of ATF4 expression under conditions in which the proteasome pathway is inhibited.

 

Materials and Methods

Cell Culture and Reagents

The media used for culture of HeLa, HT1080, H460, and HCT116 cell lines were Minimum Essential Medium (Cat. No. LM 007-7; Welgene Inc, Korea), Dulbecco’s Modified Eagle’s Medium (Cat. LM001-05; Welgene), RPMI 1640 Medium (Cat. No. LM 011- 01; Welgene), and McCoy’s 5A Medium (Cat. No. LM 005-01; Welgene), respectively. The cell lines were cultured in the media supplemented with 10% fetal bovine serum (Cat. No. 43640; JRS, USA) and 1% penicillin/streptomycin (Cat. No. 15140; GIBCO, USA) at 37℃ in a humidified atmosphere containing 5% CO2 (v/v). The reagents used were resveratrol (Cat. No. R5010; Sigma, USA) and MG132 (Cat. No. C2211; Sigma).

Gene Silencing with RNA Interference

For transient silencing experiments, cells were transfected with either negative control siRNA (Cat. No. 12935-300; Invitrogen, USA) or SIRT1 siRNA using Lipofectamine RNAiMAX (Cat. No. 13778-150; Invitrogen), according to the manufacturer’s instructions. The SIRT1 siRNA oligonucleotide was 5’-ACUUUGCUGUAACCCUGUA-3’. To obtain cell lines expressing SIRT1, cells were transfected with MFG-puro-SIRT1 plasmid, and puromycin-resistant cells were collected.

Western Blot Analysis

HeLa cells were lysed in TNN buffer (120 mM NaCl, 40 mM Tris-HCl, pH 8.0, 0.5% NP-40, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 100 mM sodium fluoride, and 2.5 μg/ml leupeptin, aprotinin, and pepstatin). Lysate samples for western blot analysis were separated using SDS-polyacrylamide gel electrophoresis. Proteins on the gel were electrophoretically transferred to nitrocellulose membranes, followed by immunoblotting with antibodies against ATF4 (Cat. No. sc-200; Santa Cruz Biotechnology), SIRT1 (Cat. No. sc-55404; Santa Cruz Biotechnology), or β-actin (Cat. No. sc-47778; Santa Cruz Biotechnology). Each protein band was detected using a luminal reagent (Cat. No. sc- 2048; Santa Cruz Biotechnology).

Total RNA Isolation and Reverse Transcriptase-Polymerase Chain Reaction

Total RNA was extracted from cells using the RNeasy Mini kit (Cat. No. 74106; Qiagen, Germany) according to the manufacturer’s protocol. For cDNA synthesis, total RNA was reverse-transcribed using the iScript cDNA synthesis kit (Cat. No. 170-8890; Bio-Rad, USA). Reverse transcriptase polymerase chain reaction (RT-PCR) was performed using the Maxime PCR PreMix kit (i-StarTaq) (Cat. No. #25167; Intron Biotechnology, Korea). The primer sequences used were as follows: ATF4 forward primer (5’-TCA AAC CTC ATG GGT TCT CC-3’) and reverse primer (5’-GGG CTC ATA CAG ATG CCA CT-3’), SIRT1 forward primer (5’-CAA ACT TTG CTG TAA CCC TGT-3’) and reverse primer (5’-CAG CCA CTG AAG TTC TTT CAT-3’), and β-actin forward primer (5’-AAG GAT TCC TAT GTC GGC) and reverse primer (5’-CAT CTC TTG CTC GAA GTC-3’).

 

Results

SIRT1 Overexpression Leads to Suppression of the ATF4 Protein Level During Proteasome Inhibition

To ascertain whether SIRT1 regulates ATF4 in association with the proteasome pathway, we examined the effect of SIRT1 on cellular ATF4 protein levels under conditionswhere the proteasome pathway is inhibited. To this end, we employed MG132, an inhibitor of 26S proteasomal degradation, which induces accumulation of unfolded protein and ER stress [22,23]. An initial experiment was performed with established HeLa cells stably transfected with a plasmid overexpressing exogenous SIRT1 [12]. Consistent with the earlier finding that ATF4 protein is short-lived and normally degraded by β-TRCP E3 ligase [15], we observed an extremely low basal ATF4 level (Fig. 1A). However, the minimal detectable ATF4 protein level was increased in the presence of MG132. Comparison of the ATF4 protein levels in the presence of MG132 between HeLa cells with and without SIRT1 transfection revealed that SIRT1 induces a decrease in the amount of ATF4 protein. The ATF4 protein level was significantly lower in cells transfected with SRT1 than those with empty vector. Under conditions where the MG132 concentration was increased from 5 to 20 μM, resulting in higher accumulation of ATF4, SIRT1-mediated downregulation of ATF4 protein was consistently observed. To further evaluate this finding, we examined ATF4 protein expression in HeLa cells transiently transfected with Myc-tagged SIRT1. In the presence of MG132, increasing the amount of transfected Myc-tagged SIRT1 led to a decrease in the ATF4 protein level (Fig. 1B). Based on these results, we conclude that SIRT1 suppresses the synthesis of ATF4 protein, which is detectable during proteasome inhibition.

Fig. 1.ATF protein synthesis is decreased upon SIRT1 overexpression. (A) The amounts of ATF4 protein were determined in the presence or absence of MG132 in HeLa cells with and without SIRT1 overexpression, established via transfection of SIRT1 (S) and empty vector (V), respectively. (B) The amount of ATF4 protein in the presence of MG132 was determined in HeLa cells transiently transfected with increasing concentrations of Myc-tagged SIRT1 plasmid (+, 1 μg; ++, 2 μg; +++, 3 μg).

Fig. 2.ATF4 protein synthesis is increased upon SIRT1 depletion. SIRT1 siRNA was transiently transfected into established HeLa cells previously transfected with SIRT1 (HeLa S) or empty vector (HeLa V), and the amount of ATF4 protein was determined in the presence or absence of MG132.

SIRT1 Depletion Induces ATF4 Protein Expression During Proteasome Inhibition

To further assess the SIRT1 suppression of ATF4 protein synthesis observed during proteasome inhibition, we examined the effect of SIRT1 depletion on the ATF4 protein level. Consistent with the data obtained with HeLa cells overexpressing SIRT1, SIRT1 depletion with siRNA in HeLa cells led to a significant increase in ATF4 protein in the presence of MG132 (lane 3 vs 4, Fig. 2). The effect of SIRT1 depletion on the ATF4 protein level was additionally observed in HeLa cells overexpressing SIRT1. Notably, ATF4 protein expression suppressed by SIRT1 overexpression (lane 3 vs 7, Fig. 2) was conversely increased with SIRT1 depletion (lane 7 vs 8, Fig. 2). However, in the absence of MG132, SIRT1 depletion did not affect the ATF4 protein level in HeLa cells transfected with either empty vector or SIRT1 (lane 1 vs 2 or lane 5 vs 6, respectively, Fig. 2). Similarly, SIRT1 overexpression had no impact on the ATF4 protein level in the absence of MG132 (lanes 1 vs 2, Fig. 1A). These findings further support SIRT1 suppression of ATF4 protein synthesis during proteasome inhibition.

SIRT1 Regulation of ATF4 Synthesis is Achieved at the Post-Transcriptional Level

Next, we focused on determining whether SIRT1 regulation of the ATF4 protein is mediated by transcriptional control of the gene. In contrast to the increased ATF4 protein expression observed in the presence of MG132, the ATF4 mRNA level was not altered upon SIRT1 depletion (Figs. 3A and 3B). The MG132 concentration did not affect ATF4 transcription (2 and 5 μM in Figs. 3A and 3B, respectively). Moreover, SIRT1 did not affect the ATF4 mRNA level in the absence of MG132 (Figs. 3A and 3B). The changes detected at the protein, but not mRNA level, in the presence of MG132 indicate that SIRT1 regulates ATF4 protein synthesis via a post-transcriptional mechanism.

Fig. 3.ATF4 protein synthesis, but not mRNA, is increased upon SIRT1 depletion. Levels of ATF4 protein and mRNA were determined with western blot assay and semiquantitative RT-PCR, respectively, in HeLa cell lines in the presence of 2 μM (A) and 5 μM (B) MG132.

Fig. 4.SIRT1 suppression of ATF4 synthesis is generally found in human cell lines. SIRT1 siRNA was transiently transfected into HT1080 (A), H460 (B), and HCT116 cell lines (C) in the presence (+) or absence (-) of MG132. ATF4 protein level was determined by western blot assay.

SIRT1 Suppression of ATF4 Synthesis is Found in Various Human Cells

As shown above, the SIRT1 suppression of ATF4 synthesis was observed in HeLa cervical carcinoma cells. To further determine whether our present finding generally occurs in human cells, we examined it using three more cell lines, including HT1080 human fibrosarcoma cells, H460 human lung carcinoma cells, and HCT116 human colon carcinoma cells. Similar to the result shown in HeLa cells, all these cell lines examined, HT1080 (Fig. 4A), H460 (Fig. 4B), and HCT116 cells (Fig. 4C), showed that SIRT1 depletion increased ATF4 synthesis upon treatment of MG132. Consistently, this was not observed in these cell lines under condition without MG132 (Figs. 4A-4C). Thus, the SIRT1 suppression of ATF4 synthesis during proteasome inhibition is generally found in human cells, not specific to a certain cell type.

 

Discussion

In the present study, we identified SIRT1 as a novel negative regulator of ATF4 protein synthesis, which is detected during proteasome inhibition. Moreover, SIRT1 suppresses ATF4 synthesis under post-transcriptional control. This conclusion was derived from the following findings: (1) ATF4 protein accumulation in the presence of MG132, a proteasome inhibitor, was reduced upon SIRT1 overexpression, but enhanced by SIRT1 depletion; (2) in the absence of MG132, SIRT1 did not affect the ATF4 protein level; and (3) the ATF4 mRNA level was not affected by SIRT1, irrespective of the presence or absence of MG132.

MG132 inhibits the proteasomal degradation of unfolded proteins, in turn promoting the unfolded protein response and ER stress [22,23]. Therefore, SIRT1 may be involved in ATF4 synthesis when proteasomal degradation of unfolded and misfolded proteins is inhibited and ER stress increasingly develops owing to the accumulation of inappropriate proteins. The finding that SIRT1 does not affect the ATF4 mRNA level indicates that SIRT1-mediated regulation is not associated with transcriptional activation of the gene. SIRT1 may thus regulate ATF4 protein synthesis at the level of translation or post-translational degradation. Recent studies have shown that SIRT1 interacts with and suppresses the phosphorylation of eIF2α [9], raising the possibility that SIRT1-mediated regulation of ATF4 protein synthesis occurs via eIF2α phosphorylation. Phosphorylation of eIF2α attenuates the inhibitory action of the open reading frame on the 5’-untranslated region located upstream of ATF4 mRNA, leading to protein translation [4,16,28]. Several kinases, including PERK, GCN2, PKR, and HRI, are responsible for eIF2α phosphorylation. Under stress conditions where these kinases are activated, SIRT1 may inhibit their activities, leading to decreased eIF2α phosphorylation and subsequent ATF4 translation. Recently, a study closely related to ours reported that ATF4 upregulates SIRT1 expression through direct binding to the SIRT1 promoter [29]. These data, taken together with our presented data, provide evidence that SIRT1 regulates ATF expression in a post-transcriptional negative feedback loop.

The ATF4 protein is stabilized via dissociation from β-TRCP E3 ligase, and affinity between the two proteins is also dependent on the phosphorylation of ATF4. RSK2, CK1, and CK2 are the suggested kinases responsible for phosphorylation of ATF4 protein [3,7,30,31]. P300 inhibits the interactions between ATF4 and β-TRCP E3 ligase, leading to dissociation of the two proteins [14]. Since SIRT1 counteracts p300 and regulates β-TRCP (unpublished data), we cannot rule out the possibility that SIRT1 participates in ATF4 protein stability. The facts that, under stress condition, SIRT1 controls nuclear translocation of transcription modulators including GAPDH [13] and ATF4 is translocated into the nucleus [27] further suggest the association of SIRT1 with ATF4. Thus, further efforts are required to solve the mechanisms underlying the modulation of cellular levels of ATF4 protein.