ER stress-linked autophagy stabilizes apoptosis effector PERP and triggers its co-localization with SERCA2b at ER-plasma membrane junctions

Abstract

Specific molecular interactions that underpin the switch between ER stress-triggered autophagy-mediated cellular repair and cellular death by apoptosis are not characterized. This study reports the unexpected interaction elicited by ER stress between the plasma membrane (PM)-localized apoptosis effector PERP and the ER Ca²⁺ pump SERCA2b. We show that the p53 effector PERP, which specifically induces apoptosis when expressed above a threshold level, has a heterogeneous distribution across the PM of un-stressed cells and is actively turned over by the lysosome. PERP is upregulated following sustained starvation-induced autophagy, which precedes the onset of apoptosis indicating that PERP protein levels are controlled by a lysosomal pathway that is sensitive to cellular physiological state. Furthermore, ER stress stabilizes PERP at the PM and induces its increasing co-localization with SERCA2b at ER-PM junctions. The findings highlight a novel crosstalk between pro-survival autophagy and pro-death apoptosis pathways and identify, for the first time, accumulation of an apoptosis effector to ER-PM junctions in response to ER stress.

Keywords: Apoptosis; Lysosomes; Macroautophagy.

Fig. 1. PERP interacts with SERCA2b at ER–PM junctions.

a Protein interacting partners of Halo-PERP were isolated from Mel202 cells using the HaloTag Mammalian Pull-Down System and identified by mass spectrometry. Maxquant intensity values for SERCA2 and PERP in three independent experiments are shown. b The interaction of PERP and SERCA2b was confirmed in two independent HaloTag pull-down experiments by immunoblotting as described in the Methods (proteins eluted from the resin by TEV enzymatic cleavage followed by a successive SDS-based elution). Figure shows two different exposures of SERCA2b panel to allow visualization of lower intensity bands. c PERP–SERCA2 complex formation was validated by SERCA2 IP in Mel202 cells expressing HaloTag or Halo-PERP using a HaloTag antibody. d Diagram of SERCA2a–c proteins, indicating the relative positions of the peptide identified by mass spectrometry (black) and the antibody immunogen sites (grey) used for validation of the SERCA2b–PERP interaction. Antibody 1 was used in HaloTag pull down validation (b) and antibody 2 was used for IP of SERCA2 (c). e Super-resolution images of HeLa BAC Venus-PERP cells co-expressing mCherry-SERCA2b, junctions between the ER and PM shown by arrows. Scale bar: 20 µm in full image and 5 µm in zoom panel

Fig. 2. PERP protein is stabilized by ER stress, independent of p53 transcriptional regulation.

HCT116 cells were treated with 1 μg/ml BFA for the indicated time points and changes to the levels of a SERCA2b mRNA (n = 4, one-way ANOVA, F = 3.442, p = 0.0239*), b SERCA2b protein (n = 7), c PERP protein (n = 7, one-way ANOVA, F = 3.025, p = 0.0329*) and d PERP mRNA (n = 4, one-way ANOVA, F = 5.297, p = 0.0036**) were detected by RT-PCR or Western blot and normalized to the level of GAPDH. e HCT116 p53−/− cells were treated with 1 µg/ml BFA and the response of PERP protein levels was detected by immunoblotting

Fig. 3. PERP accumulates at plasma membrane puncta during ER stress.

a, b Super-resolution images of HeLa BAC Venus-PERP cells at the z-stack centre (a) and base of the cell (b), either non-treated (NT) or BFA-treated (16 h). Scale bars: 20 µm

Fig. 4. PERP protein is actively degraded by the lysosome.

a HCT116 cells were treated with 30 μg/ml CHX for the indicated time points, and the protein levels of PERP and p53 were detected by immunoblotting. Histogram shows PERP protein levels normalized to GAPDH. One-way ANOVA, n = 3, F = 12.63, p = 0.0003***. b HCT116 cells were treated with 20 μM MG132 or 100 μM CQ for the indicated time points and PERP protein levels were detected by Western blot. Graph represents PERP protein levels normalized to GAPDH. One-way ANOVA, n = 3, MG132 PERP: F = 0.8849, p = 0.5204 ns; CQ PERP: F = 13.03, p = 0.0002***. c HCT116 cells were treated with 20 μM MG132 and the protein levels of p53 were detected by immunoblotting and normalized to the level of GAPDH. One-way ANOVA, n = 3, F = 11.58, p = 0.0003***. d HCT116 cells were treated with 1, 10 and 100 nM of Baf and the protein levels of LC3B and PERP were detected 24 h post-treatment by immunoblotting

Fig. 5. PERP is selectively upregulated in response to sustained autophagy induction, correlating with apoptosis.

a Protein levels of autophagy marker LC3B in HCT116 cells treated with 1 μg/ml BFA. b Autophagy flux analysis of p62 levels in HCT116 cells treated with 1 µg/ml BFA in the presence and absence of 10 mM Baf. Protein levels quantified by densitometry and normalized to the level of GAPDH. Two-way ANOVA with Bonferroni post hoc test, n = 3. c HCT116 cells were cultured in the presence (+) or absence (−) of serum for 24 h and p62 and PERP protein levels were detected and normalized to the level of GAPDH. Blot from one independent experiment shown, performed in technical triplicate. Student’s t-test, n = 3, p62: p = 0.0361*, PERP: p = 0.02*. d HCT116 cells were serum starved for the indicated time points and PERP protein levels were detected by Western blot and normalized to the level of GAPDH. One-way ANOVA, n = 5, F = 7.426, p = 0.0003***. e HCT116 cells were serum-starved for the indicated time points and apoptosis induction was measured by flow cytometry using an Alexa Fluor 647 annexin V conjugate. One-way ANOVA, n = 4, F = 41.00, p < 0.0001****

Fig. 6. SERCA2b expression correlates with PERP protein stabilization and apoptosis induction.

a mCherry (24 h) or mCherry-SERCA2b were expressed in Mel202 cells and PERP protein levels were detected by Western blot. b Graph shows PERP mRNA and protein levels normalized to the level of GAPDH in Mel202 cells expressing mCherry and mCherry-SERCA2b. Student’s t-test, n = 3, 48 h PT: p = 0.005**; 72 h PT: p = 0.0017**. c The percentage of apoptotic Mel202 cells expressing eGFP and eGFP-SERCA2b was measured at 24 and 48 h post-transfection determined by flow cytometry using Alexa Fluor 647 annexin V. Student’s t-test, n = 3, 24 h PT: p = 0.4924 ns, 48 h PT: p = 0.0048**. d Super-resolution images of HeLa BAC Venus-PERP cells expressing mCherry-SERCA2b for 24 and 48 h; scale bars: 20 µm

Fig. 7. PERP increasingly co-localizes with SERCA2b during ER stress.

a HeLa BAC Venus-PERP cells co-expressing mCherry-SERCA2b were treated with 1 μg/ml BFA for 16 h and co-localization was assessed at the mid and bottom optical sections. Scale bars: 20 µm. b Enlargement of boxed areas with rainbow pseudo colour-coded distribution of Venus-PERP. Scale bars: 5 µm