p53-Dependent PUMA to DRAM antagonistic interplay as a key molecular switch in cell-fate decision in normal/high glucose conditions

Abstract

Background: As an important cellular stress sensor phosphoprotein p53 can trigger cell cycle arrest and apoptosis and regulate autophagy. The p53 activity mainly depends on its transactivating function, however, how p53 can select one or another biological outcome is still a matter of profound studies. Our previous findings indicate that switching cancer cells in high glucose (HG) impairs p53 apoptotic function and the transcription of target gene PUMA.

Methods and results: Here we report that, in response to drug adriamycin (ADR) in HG, p53 efficiently induced the expression of DRAM (damage-regulated autophagy modulator), a p53 target gene and a stress-induced regulator of autophagy. We found that ADR treatment of cancer cells in HG increased autophagy, as displayed by greater LC3II accumulation and p62 degradation compared to ADR-treated cells in low glucose. The increased autophagy in HG was in part dependent on p53-induced DRAM; indeed DRAM knockdown with specific siRNA reversed the expression of the autophagic markers in HG. A similar outcome was achieved by inhibiting p53 transcriptional activity with pifithrin-α. DRAM knockdown restored the ADR-induced cell death in HG to the levels obtained in low glucose. A similar outcome was achieved by inhibition of autophagy with cloroquine (CQ) or with silencing of autophagy gene ATG5. DRAM knockdown or inhibition of autophagy were both able to re-induce PUMA transcription in response to ADR, underlining a reciprocal interplay between PUMA to DRAM to unbalance p53 apoptotic activity in HG. Xenograft tumors transplanted in normoglycemic mice displayed growth delay after ADR treatment compared to those transplanted in diabetics mice and such different in vivo response correlated with PUMA to DRAM gene expression.

Conclusions: Altogether, these findings suggest that in normal/high glucose condition a mutual unbalance between p53-dependent apoptosis (PUMA) and autophagy (DRAM) gene occurred, modifying the ADR-induced cancer cell death in HG both in vitro and in vivo.

Keywords: Autophagy; Cancer; Chemotherapy; DRAM; Diabetes; Hyperglycemia; PUMA; p53.

Fig. 1

High glucose (HG) switched the adriamycin (ADR)-induced p53 transcriptional activity from PUMA to DRAM. (a) RKO and HCT116 cells were kept in low glucose (LG) or high glucose (HG) medium for 24 h and then treated with ADR (2 μg/ml) for 16 h before being assayed for semi-quantitative RT-PCR analysis of PUMA and DRAM gene expression. 28S was used as a control for efficiency of RNA extraction and transcription. Histograms representing quantification of PUMA or DRAM/28S ratio as assessed by densitometric analysis are shown. Densitometric values were quantified using the ImageJ software and normalized to control. The values of control were set to 1. The data are presented as means ± S.D. of three independent experiments. *P < 0.001. (b) RKO and HCT116 cells were treated with ADR (2 μg/ml) for 16 h in LG and HG medium, with or without p53 inhibitor pifithrin-α (PFT-α) (30 μM) before being assayed for semi-quantitative RT-PCR analysis of PUMA and DRAM gene expression. 28S was used as a control for efficiency of RNA extraction and transcription. One representative experiment is shown. (c) HCT116-p53^−/− cells were treated and assayed as in (a)

Fig. 2

HG increased autophagy in ADR-treated cells. (a) RKO stably transfected with GFP-LC3 plasmid were kept in low glucose (LG) or high glucose (HG) medium for 24 h and then treated with ADR (2 μg/ml) for 16, in the presence or absence of autophagy inhibitor chloroquine (CQ, 25 μM) for 4 h before observation to count GFP-LC3 puncta under fluorescence microscopy. Green indicates GFP-LC3. One of 10 representative micrographs is shown. (b) The relative number of GFP-LC3-positive cells was calculated from 10 random fields. The data are presented as the means ± S.D. from three independent experiments. (c) RKO cells were kept in low glucose (LG) or high glucose (HG) medium for 24 h and then treated with ADR (2 μg/ml) for 16 h in the presence or absence of 25 μM chloroquine (CQ), and the expression of LC3-I/II was measured by western blot analysis. One representative experiment is shown. β-actin was used as internal control. (d) Findings as in (c) were assessed by quantitative analysis of LC3II-I/β-actin protein levels and shown as histograms. The data are presented as the means ± S.D. from three independent experiments. *P < 0.001

Fig. 3

Increase of autophagy of ADR-treated cells in HG was in part depended on p53 activity. (a) HCT116 cells were treated with ADR (2 μg/ml) for 16 h in LG and HG medium, with or without p53 inhibitor pifithrin-α (PFT-α) (30 μM) before being assayed for western blot analysis. Densitometric values of LC3II/I/β-actin protein levels were quantified using the ImageJ software and normalized to control. One representative experiment is shown. (b) HCT116 cells were transfected with ctr-siRNA and siDRAM and 36 h after transfection DRAM expression was assessed by RT-PCR analysis. One representative experiment is shown. (c) HCT116 cells, transfected with ctr-siRNA and siDRAM, were kept in HG medium for 24 h and then treated with ADR (2 μg/ml) for 16 h before the expression of p62 was measured by western blot analysis. Densitometric values of p62/β-actin protein levels were quantified using the ImageJ software and normalized to control and the results are shown under the images. One representative experiment is shown. Anti β-actin was used as protein loading control

Fig. 4

Autophagy inhibition or DRAM silencing restored ADR-induced cell death in HG. (a) HCT116 cells were treated with ADR (2 μg/ml for 24 h) in low (−) and high glucose (HG) condition with or without 25 μM CQ (for 16 h). After treatments, cells were in part fixed and stained with propidium iodide (PI) for subG1 evaluation (upper panel) or lysed and analyzed by western immunoblotting to assess PARP cleavage (lower panel); relative quantification of PARP cleavage/β-actin ratio is shown. One representative experiment is shown. Anti-β-actin was used as protein loading control. *P < 0.001. (b) RKO and HCT116 cells, transfected with ctr-siRNA and siDRAM or left untransfected, were kept in low glucose (−) or high glucose (HG) medium for 24 h and then treated with ADR (2 μg/ml) for 24 h before the percentage of dead cells was scored by trypan blue exclusion. The data are presented as the means ± S.D. from three independent experiments. *P < 0.001. (c) RKO cells, transfected with ctr-siRNA and with siATG5 or left untransfected, were kept in low glucose (−) or high glucose (HG) medium for 24 h and then treated with ADR (2 μg/ml) for 24 h before the percentage of dead cells was scored by trypan blue exclusion. The data are presented as the means ± S.D. from three independent experiments. *P < 0.001

Fig. 5

Autophagy inhibition or DRAM silencing restored ADR-induced PUMA transcription in HG. (a) RKO and HCT116 cells were kept in low glucose (LG) or high glucose (HG) medium for 24 h and then treated with ADR (2 μg/ml) for 16 h in the presence or absence of 25 μM CQ, before being assayed for semi-quantitative RT-PCR analysis of PUMA gene expression. 28S was used as a control for efficiency of RNA extraction and transcription. One representative experiment is shown. (b) RKO and HCT116 cells, transfected with with siRNA and siDRAM or untransfected, were kept in low glucose (−) or high glucose (HG) medium for 24 h and then treated with ADR (2 μg/ml) for 16 h before being assayed as in (a). One representative experiment is shown. (c) Densitometric values of three independent experiments as in (b) were quantified using the ImageJ software and normalized to control. The values of control were set to 1. The data are presented as fold of induction ± S.D. *P < 0.005

Fig. 6

Diabetes reduced the effect of chemotherapy that correlated with increased DRAM and reduced PUMA gene expression, in vivo. (a) Blood glucose concentration was evaluated in normoglycemic (Normo) and streptozotocin (SZT)-treated mice (diabetic). Data are presented as means ± S.D. *P < 0.005. (b) After mice reached a glucose concentration exceeding 300 mg/dl (considered diabetic) upon SZT treatment, solid tumors were obtained by injecting i.m. RKO cells on the flank of each mouse. ADR treatment was performed when the tumors became palpable. Ten days after ADR treatment, the size of tumors showed statistical significant growth delay in normoglycemic versus SZT mice. The data are presented as fold reduction ± S.D. *P < 0.005. (c) Tumors measured in (b) were then explanted from normoglycemic (Normo) and streptozotocin (SZT)-treated mice and total mRNA was analysed by RT-PCR of PUMA and DRAM gene expression. 28S was used as a control for efficiency of RNA extraction and transcription. (c) Densitometric analysis of gene expression in (b) was plotted as expression ratio to 28S. *P < 0.001