PARP-1 modulation of mTOR signaling in response to a DNA alkylating agent

Abstract

Poly(ADP-ribose) polymerase-1 (PARP-1) is widely involved in cell death responses. Depending on the degree of injury and on cell type, PARP activation may lead to autophagy, apoptosis or necrosis. In HEK293 cells exposed to the alkylating agent N-methyl-N'-nitro-N'-nitrosoguanine (MNNG), we show that PARP-1 activation triggers a necrotic cell death response. The massive poly(ADP-ribose) (PAR) synthesis following PARP-1 activation leads to the modulation of mTORC1 pathway. Shortly after MNNG exposure, NAD⁺ and ATP levels decrease, while AMP levels drastically increase. We characterized at the molecular level the consequences of these altered nucleotide levels. First, AMP-activated protein kinase (AMPK) is activated and the mTORC1 pathway is inhibited by the phosphorylation of Raptor, in an attempt to preserve cellular energy. Phosphorylation of the mTORC1 target S6 is decreased as well as the phosphorylation of the mTORC2 component Rictor on Thr1135. Finally, Akt phosphorylation on Ser473 is lost and then, cell death by necrosis occurs. Inhibition of PARP-1 with the potent PARP inhibitor AG14361 prevents all of these events. Moreover, the antioxidant N-acetyl-L-cysteine (NAC) can also abrogate all the signaling events caused by MNNG exposure suggesting that reactive oxygen species (ROS) production is involved in PARP-1 activation and modulation of mTOR signaling. In this study, we show that PARP-1 activation and PAR synthesis affect the energetic status of cells, inhibit the mTORC1 signaling pathway and possibly modulate the mTORC2 complex affecting cell fate. These results provide new evidence that cell death by necrosis is orchestrated by the balance between several signaling pathways, and that PARP-1 and PAR take part in these events.

Figure 1. PARP-1 activation following MNNG exposure affects PAR, NAD⁺, AMP, and ATP levels.

(a) HEK293 cells were treated with 100 µM MNNG alone (left panel) or in combination with 5 µM of the PARP inhibitor AG14361 (right panel) and monitored for 8 hours. PAR accumulation was detected by western blot using 96-10 antibody. The blots were also probed with actin antibody to show equal loading. (b) NAD⁺ and AMP levels and (c) ATP levels were measured in HEK293 cells treated with 100 µM MNNG alone or in combination with 5 µM AG14361. (d) AMP/ATP ratios were calculated from AMP and ATP values at each time point. NAD⁺, AMP, and ATP levels were measured as described in Materials and Methods. Data are presented as the mean ± standard error of the mean (SEM) of four independent experiments.

Figure 2. PARP-1 activation is associated with AMPK activation and modulation of mTORC1/2 signaling.

HEK293 cells were treated with 100 µM MNNG alone or in combination with 5 µM AG14361. Cell lysates were analyzed by immunoblotting for (a) AMPK phosphorylation on Thr172, (b) Raptor phosphorylation on Ser792, (c) S6 phosphorylation on Ser240/244, (d) Rictor phosphorylation on Thr1135, and (e) Akt phosphorylation on Ser473. Total protein (AMPK, Raptor, S6, Rictor, and Akt) was measured as an internal control (lower panels). Immunoblots are representative of three to four independent experiments.

Figure 3. Effect of PARP-1 activation on mTORC1 signaling in HeLa cells.

Cells were treated with 100 µM MNNG and monitored for up to 8 hours. (a) PAR accumulation was detected by western blot using 96-10 antibody. The blots were also probed with actin antibody to show equal loading. (b) AMP levels. (c) AMPK phosphorylation on Thr172 (upper panel) and total AMPK protein (lower panel). (d) Raptor phosphorylation on Ser792 (upper panel) and total Raptor protein (lower panel).

Figure 4. Effect of MNNG exposure on Beclin 1 expression and on the autophagic vesicles marker LC3.

HEK293 cells were treated with 100 µM MNNG alone or in combination with 5 µM AG14361 one hour prior to MNNG exposure. (a) Beclin 1 expression and the conversion of LC3-I to LC3-II were detected by immunoblotting. The blots were also probed with actin antibody to show equal loading. (b) Cells were transiently transfected with the GFP-LC3 plasmid for 24 hours and incubated with MNNG alone or in combination with AG14361. Nuclei were stained with Hoechst 33342. The distribution of GFP-LC3 was examined by fluorescence microscopy 4 hours after the start of MNNG exposure and representative cells were photographed. As a positive control for GFP-LC3 redistribution, cells were treated with 1 mM H₂O₂ for 48 hours (lower right panel). Punctuated distribution of GFP-LC3 is indicated by the arrows. (c) HEK293 cells were treated with MNNG alone or in combination with the lysosomal protease inhibitor hydrochloroquine sulfate (HCQ) and LC3-II accumulation was detected by immunoblotting. (d) LC3-II/Actin ratios were calculated at each time point. Data are presented as the mean ± standard error of the mean (SEM) of three independent experiments. * P<0.005. One way analysis of variance (ANOVA) shows no significant changes in LC3-II accumulation in MNNG+HCQ-treated cells between 0 and 4 hours.

Figure 5. Effect of the antioxidant N-acetyl L-cysteine (NAC) on PARP-1 activation and mTORC1/2 signaling.

HEK293 cells were treated with 100 µM MNNG alone or in combination with 10 mM NAC. Cell lysates were analyzed by immunoblotting for (a) PAR accumulation using 96-10 antibody. The blots were also probed with actin antibody to show equal loading. (b) AMPK phosphorylation on Thr172, (c) Raptor phosphorylation on Ser792, (d) S6 phosphorylation on Ser240/244, (e) Rictor phosphorylation on Thr1135, and (f) Akt phosphorylation on Ser473. Total protein (AMPK, Raptor, S6, Rictor, and Akt) was measured as an internal control (lower panels). Immunoblots are representative of three independent experiments.

Figure 6. Effect of PARP-1 activation on MNNG-induced cell death.

HEK293 cells were treated with 100 µM MNNG alone or in combination with 5 µM AG14361 or 10 mM NAC one hour prior to MNNG exposure. Cell death was evaluated as percentage of PI positive cells 6 hours after MNNG treatment with or without (a) AG14361 or (b) NAC by staining with PI and Annexin-V-FITC coupled with flow cytometry. Data are presented as the mean ± SEM of at least three independent experiments. (c) Phase-contrast images of control cells (upper left panel) and 6 hours after the start of MNNG exposure (lower left panel). HEK293 cells treated with the PARP inhibitor alone (upper right panel) or before MNNG exposure (MNNG + AG, lower right panel) were protected from detachment and were morphologically similar to control cells. (d) Representative data showing cell distribution as evaluated by staining with PI (y axis) and Annexin-V-FITC (x axis) from a population of 10,000 cells.

Figure 7. Evaluation of the type of cell death following MNNG exposure (a) Cell distribution at different time points following MNNG exposure by staining with PI and Annexin-V-FITC coupled with flow cytometry.

(b) Phase-contrast images of cells at different time points following MNNG exposure. (c) Immunoblot analysis of PARP-1. As a positive control for PARP-1 cleavage, HEK293 cells were treated with 100 µM VP-16 for 24 hours. Actin was used as an internal control.

Figure 8. Effect of genetic inhibition of AMPK on PAR synthesis and cell death following MNNG exposure.

Cells were transfected with AMPα1 and α2 siRNAs or with control siRNA for 96 h and then treated with 100 µM MNNG. (a) The knock-down of AMPKα1 and AMPKα2 was determined by western blotting. Actin expression was measured as an internal control. (b) PAR accumulation in MNNG treated cells with or without AMPK knock-down was detected by western blot using 96-10 antibody. The blots were also probed with actin antibody to show equal loading. (c) Cell death was evaluated 6 hours after MNNG treatment by staining with PI and Annexin-V-FITC coupled with flow cytometry. Data are presented as the mean ± SEM of three independent experiments.

Figure 9. Effect of AMPK knock-down on mTORC1/2 signaling following PARP-1 activation by MNNG exposure.

HEK293 cells were treated with siRNA directed against α1 and α2 subunits of AMPK (right panel) or with control siRNA (left panel). Following 100 µM MNNG exposure, cell lysates were analyzed by immunoblotting for (a) AMPK phosphorylation on Thr172, (b) Raptor phosphorylation on Ser792, (c) S6 phosphorylation on Ser240/244, (d) Rictor phosphorylation on Thr1135, and (e) Akt phosphorylation on Ser473. Total protein (Actin, Raptor, S6, Rictor, and Akt) was measured as an internal control (lower panels). In (a), total AMPK was also measured to show the reduced AMPK levels following AMPK knockdown. Immunoblots are representative of three independent experiments.

Figure 10. Schematic representation of the possible effects of PARP-1 activation on mTORC1 and mTORC2 components and cell death.

Following MNNG exposure, reactive oxygen species (ROS) are produced and DNA double-strand breaks (DSB) are formed. PAR transiently accumulates and its metabolism results in a decrease in ATP and NAD⁺ levels, while cellular AMP increases concomitantly. The high AMP levels activate AMPK by LKB1. AMPK activation causes inactivation of the mTORC1 pathway through phosphorylation of Raptor and inhibition of S6 phosphorylation. S6 phosphorylation may also be affected by the reduced ERK1/2 phosphorylation observed following PAR synthesis. Even though mTORC1 inhibition is the first step towards the establishment of an autophagic state, autophagy is not activated. Rictor phosphorylation is affected by PAR synthesis and Akt phosphorylation at Ser473 is impaired. Finally, in HEK293 cells, PARP-1 activation by MNNG exposure results in necrotic cell death. The PARP inhibitor AG14361 and the antioxidant N-acetyl-L-cysteine (NAC) inhibit PAR synthesis and prevent cell death.