p53‐dependent transcriptional suppression of BAG3 protects cells against metabolic stress via facilitation of p53 accumulation

Abstract Solid tumour frequently undergoes metabolic stress during tumour development because of inadequate blood supply and the high nutrient expenditure. p53 is activated by glucose limitation and maintains cell survival via triggering metabolic checkpoint. However, the exact downstream contributors are not completely identified. BAG3 is a cochaperone with multiple cellular functions and is implicated in metabolic reprogramming of pancreatic cancer cells. The current study demonstrated that glucose limitation transcriptionally suppressed BAG3 expression in a p53‐dependent manner. Importantly, hinderance of its down‐regulation compromised cellular adaptation to metabolic stress triggered by glucose insufficiency, supporting that BAG3 might be one of p53 downstream contributors for cellular adaptation to metabolic stress. Our data showed that ectopic BAG3 expression suppressed p53 accumulation via direct interaction under metabolic stress. Thereby, the current study highlights the significance of p53‐mediated BAG3 suppression in cellular adaptation to metabolic stress via facilitating p53 accumulation.

BAG3 is one member of the human BAG cochaperone family (BAG1-6) that interacts with the ATPase domain of the heat shock protein 70 (HSP70) via conserved BAG domain. 7,8 BAG3 exhibits a wide range of biological functions including the regulation of stress responses, autophagy, cellular survival, apoptosis and viral replication. 9,10 BAG3 is constitutively expressed in cardiac myocytes and skeletal muscle cells, but is induced by different stimuli in many other cell types. 9,[11][12][13][14][15] Inducible BAG3 expression usually serves as a cellular protective mechanism against stressful stimuli. 7, [16][17][18][19][20][21] Recently, we have demonstrated that BAG3 promotes aerobic glycolysis and growth of pancreatic cancer cells by stabilizing HK2 mRNA via interaction with HK2 transcripts, 22 assigning BAG3 another functions to RNA binding and metabolic reprogramming.
In the present study, we show that BAG3 is suppressed in a p53-dependent pattern under metabolic stress triggered by glucose limitation. Ectopic BAG3 expression inhibits p53 accumulation via direct interaction and compromises cellular adaptation to glucose limitation. To our knowledge, the current study for the first time demonstrates that p53-mediated suppression of BAG3 facilitates p53 accumulation and cellular survival upon glucose insufficiencyinduced metabolic stress. These results significantly enhance our understanding on the molecular mechanism(s) by which cells adapt to metabolic stress. Moreover, these insights provide a novel therapeutic target to alter metabolic activity and selectively eliminate cancer cells.

| Chromatin immunoprecipitation (ChIP)
Chromatin immunoprecipitation assays were performed using a kit from Upstate Biotechnology Inc The cells were fixed with 1% formaldehyde to cross-link chromatin according to the manufacturer's instructions. The cell lysates were prepared and sonicated on ice to break chromatin DNA to an average length of 400 bp. After a preclearing step, immunoprecipitation was carried out at 4°C overnight with anti-p53 antibody or normal goat IgG (negative control antibody). Immune complexes were collected with salmon sperm DNA saturated protein A-agarose beads. After extensive washing, the immunoprecipitated complexes were eluted with 0.1 mol/L NaHCO 3 and 1% SDS. The protein-DNA cross-links were then reversed by incubating at 65°C for 5 hours. DNA was subsequently purified using a proteinase K digestion with a phenol: chloroform extraction and an ethanol precipitation. Real-time PCR was performed using PCR primers specific for the BAG3(GI: 9531) RE1 sequence between nucleotide positions −29885 and −29876 including the forward primer 5'-GTGATTCTCCCACCTCAGCCTTCT-3' and the reverse primer 5'-GGCCAATGTAGTGAAACCCCGTCT-3'. PCR amplification resulted in a 110 bp product that contained the BAG3 RE1 sequence.
In addition, PCR amplification was conducted to amplify the RE2 sequence between nucleotide positions +5143 and +5152 using the forward primer 5'-ACAGGCTAAAAGTGGGTTGGA-3' and the reverse primer 5'-TGTGGCCCAGAAGCCACTTA-3'), which generated an 85 bp product. In addition, the p21 sequence was used as a positive control by amplifying the sequence between −2421 and −2412 positions using the forward primer 5'-CTGAGCCTCCCTCCATCC-3' and the reverse primer 5'-GAGGTC TCCTGTCTCCTACCA TC-3', which generated a 189 bp amplification product. Standard curves were then calculated from the amplification data. The abundances of the BAG3 promoter that were amplified were considered as the IP and Input. In order to facilitate comparison, amplification results were expressed as IP/Input ratios of PCR products.

| Cell proliferation assay
Cells were seeded into six-well plates at a density of 100 000 cells per well, 2 mL of medium supplement 10% FBS per well. Change the medium every 24 hours. The cells were incubated with trypan blue solution at 1:1. Cell number was determined by counting with automated cell count kit.

| Detection of apoptotic cell death
To detect apoptotic cell death, cells were washed twice with PBS and then stained using the Annexin V/PI Apoptosis Detection Kit according to the manufacturer's instructions. After staining with Annexin V and PI for 30 minutes, samples were analysed using a fluorescenceactivated cell scanner flow cytometer.

| Western blot and immunoprecipitation
Cells were cultured and harvested followed by lysing in buffer containing 20 mmol/L Tris-HCl, 150 mmol/L NaCl, 2 mmol/L EDTA, 1% Triton-X100 and freshly added protease and phosphatase inhibitors.
Equivalent amounts of protein were subjected to SDS-PAGE under reducing conditions in 10% SDS-PAGE and then transferred to PVDF membranes. Cell lysates were pre-cleared with protein A/G magnetic beads for immunoprecipitation, and the protein A/G magnetic beads were treated with various antibodies followed by incubation overnight at 4°C. The immunoprecipitants were washed with lysis buffer three times and analysed by Western blot analysis, which was performed using primary antibodies against Flag, Myc antibodies.

| Crystal Violet Assay
To visualize cells with crystal violet, cells were seeded into 12-well plates at a density of 50 000 cells per well along with 1 mL of media that was supplemented with 10% FBS and 1 mmol/L glucose, followed by incubation for 16 hours. After complete washing with PBS, cells were incubated with crystal violet solution for 30 minutes at room temperature. The stain was solubilized in 33% acetic acid, and staining was evaluated at 570 nm.

| Cell Cycle Analysis
To investigate alteration of cell cycles, cells in logarithmic growth phase were seeded into six-well plates and maintained in FBS-free medium for 12 hours. The cells were then fixed in 70% ethanol over 18 hours. The fixed cells were stained with a 50 μg/mL propidium iodide (PI) solution and Rnase A for 1 hour. The fluorescence was measured using the Becton Dickinson FACScan. Cell Cycle phase analysed by ModFIT software.

| In situ proximity ligation assay
In situ proximity ligation assays (PLA) were conducted using the Duolink PLA kit (Sigma-Aldrich) according to the manufacturer's protocol. For these analyses, cells were seeded into a chamber slide at a concentration of 5000 cells per well. PLA probes were diluted in the appropriate buffer at a ratio of 1:5 and added to the sample, incubated 1 hour at 37°C. Ligation, dilute the 5x Ligation buffer 1:5 in high purity water and mix for 1 hour and wash the slides 5 minutes in 1x Wash Buffer A at room temperature. Amplification, tap off the ligation solution from the slides, add polymerase to the 1x amplification buffer at a 1:80 dilution and mix. Incubate the slides in a pre-heated humidity chamber for 100 minutes at 37°C. Fluorescence was visualized under a confocal microscopy, using at least a 60x objective.   Primary MEF cells can attach to the walls of flasks. Consequently, primary MEF cells were maintained in DMEM that was supplemented with 10% foetal bovine serum and incubated at 37°C with 5% CO 2 .

| Statistical analyses
The statistical significance of differences among data was analysed with an analysis of variance (ANOVA) and post hoc Dunnett's tests.
Statistical significance was defined as P < .05. All experiments were repeated in triplicate, and the data are expressed as the mean ± SD (standard deviation).

| Glucose insufficiency specifically decreases BAG3 expression in a p53-dependent manner
The activity of p53 is critical for cell survival upon metabolic stress induced by lack of glucose. However, the exact its downstream effectors remained largely unknown. Real-time PCR shows that BAG3 was significantly decreased by extremely low glucose (1 mmol/L) in HCT116 cells with wild-type p53 (p53+/+), while its expression was unaltered in HCT116 cells with p53 deletion (p53−/−) ( Figure 1A). Consistent with mRNA expression, Western blot demonstrated that 1mM glucose culture decreased BAG3 expression in p53+/+ HCT116 cells, while unaltered BAG3 expression in p53−/− HCT116 cells ( Figure 1B). It should be noted that compared with its control partner, p53 null HCT116 cells exhibited higher BAG3 expression levels under high glucose culture condition ( Figure 1A,B). Similar like in HCT116 cells, low glucose decreased BAG3 mRNA ( Figure 1C) and protein ( Figure 1D) in p53+/+ MEF cells. In addition, deletion of p53 increased BAG3 expression in MEF cells, which was unaltered by low glucose ( Figure 1C,D). Low glucose also decreased BAG3 mRNA ( Figure 1E) and protein ( Figure 1F) expression levels in Bel-7402 and MCF7 cells, both of which contain wild-type p53. Nascent RNA analysis demonstrated that glucose insufficiency significantly decreased novel synthesis of BAG3 mRNA in HCT116 cells with p53 ( Figure 1G). These data indicated that BAG3 might be suppressed by glucose insufficiency in a p53-dependent manner. Searching on BAG3 gene found two potential p53 responsive elements (RE1 and RE2) located on distant upstream and intron 1 of BAG3 gene, respectively ( Figure 1H). ChIP was then performed and demonstrated that p53 was recruited to both RE1 and RE2 of BAG3 gene, which was increased by glucose insufficiency ( Figure 1I). As a positive control, p53 was recruited to p21 promoter, which was augmented by glucose insufficiency ( Figure 1I). On the contrary, a fragment BAG3 intron that does not contain a potential p53 responsive element (NRE) was not immunoprecipitated by p53 ( Figure 1I). N.S.

| Glucose limitation had no effects on expression of other BAG members
Glucose insufficiency resulted in similar pattern of alteration in BAG5 mRNA ( Figure 2A). BAG1, BAG4 and BAG6 mRNA expression was unaltered, while BAG2 mRNA was increased irrespective of p53 status by glucose insufficiency (Figure 2A). Western blot demonstrated that other members of BAG family including BAG5 were unaltered by glucose insufficiency ( Figure 2B). These data indicated that BAG3 was specifically suppressed by glucose insufficiency.

| BAG3 reduction is a protective response of cells to glucose insufficiency
To investigate the possible function of BAG3 down-regulation in response to glucose insufficiency, BAG3 was ectopically introduced to HCT116 cells ( Figure 3A). Forced BAG3 expression significantly decreased viable HCT116 cell numbers irrespective of p53 status under low glucose culture ( Figure 3B). Flowcytometry  Figure 3L). In addition, glucose insufficiency did not trigger cell cycle arrest in MEF cells with BAG3 KI ( Figure 3M).

| BAG3 decreases p53 activity under glucose insufficiency
Ectopic BAG3 decreased p53 expression under both high and extremely low glucose culture in HCT 116 ( Figure 4A), MCF7 and Bel-7402 ( Figure 4B), as well as MEF ( Figure 4C) cells. Consistent with p53 expression, luciferase activity assays demonstrated that glucose insufficiency significantly increased the luciferase activity of pGL13 Luc construct, which contains p53 responsive element, in HCT116 cells ( Figure 4D). Ectopic BAG3 expression significantly decreased the luciferase activity of pGL13 Luc under glucose insufficiency ( Figure 3D). The luciferase activity of control Luc construct was significantly decreased by glucose insufficiency in control HCT116 cells, while it was rarely affected in cells with ectopic BAG3 expression ( Figure 4D). Similar regulation of control and pGL13 Luc activities by ectopic BAG3 expression was observed in MEF cells ( Figure 4E).
To explore the potential mechanisms underlying maintenance of control Luc expression under glucose insufficiency by ectopic BAG3 expression, novel protein synthesis was analysed using HPG and demonstrated that glucose insufficiency significantly blocked protein synthesis in control HCT116 ( Figure 4F) and MEF ( Figure 4G) cells. On the other hand, HCT116 ( Figure 4F) and MEF ( Figure 4G) cells with BAG3 overexpression demonstrated active protein synthesis under glucose insufficiency. BAG3 knockdown increased p53 expression in HCT116 cells ( Figure 4H). BAG3 knockdown suppressed proliferation of HCT116 cells under high glucose culture, while increased cell viability upon glucose insufficiency ( Figure 4I).

| BAG3 compromises cellular adaption to metabolic stress via direct interaction with p53
As global screen of BAG3 interactomes has identified p53 might be a potential binding partner of BAG3, we then explored whether

| D ISCUSS I ON
Sufficient glucose supply is a mandatory requirement for execution of replicative cell division. Proliferating mammalian cells sense metabolic stress induced by short of nutrient supply, once glucose provision is insufficient, cell cycle checkpoint occurs at the G1/S boundary despite the sustained availability of amino acid substrates. 5 The coordination of glucose availability with cell cycle transition permits cell to adjust to unfavourable microenvironments and represents a nutrient-sensing pathway to suspend cell growth before nutrient availability falls below threshold that cannot sustain cell survival any more. 5 Therefore, understanding signalling pathways underlying cellular adaptation to metabolic stress has emerged as a focus in the field of cancer.
In addition to its critical functions in the control of cell cycle and DNA damage response, the tumour suppressor protein p53 regulates a range of cellular metabolic procedures including glycolysis, 23,24 oxidative phosphorylation, 23 pentose phosphate pathway (PPP), 25 and glutaminolysis. 26 In addition, activation of p53 by glucose limitation plays a critical role to suppress cell cycle progression. 5 In contrast to its role in promoting apoptosis after DNA damage stress, p53 also contributes to cellular survival during metabolic stress that is induced by glucose limitation. 5 p53 is well known for its function as a transcription factor, and its activation produces versatile biological Flag-BAG3

Myc-p53
Flag-BAG3  In summary, this study demonstrates the importance of p53-mediated BAG3 suppression in protection of cells from metabolic stress induced by glucose limitation. BAG3 directly interacts with p53 to promote calpain-dependent degradation of p53, and thereby, BAG3 suppression liberates p53 and facilitates its accumulation during metabolic stress. The current study provides important insights for understanding the molecular mechanism(s) underlying the p53-mediated cellular adaptation to metabolic stress. The results from this study thus provide a potential opportunity to develop novel therapeutic strategy to get rid of cancer cells.

ACK N OWLED G EM ENTS
This work was partly supported by National Natural Science

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon request.