Figure 1.
(a) The subcellular distribution of ER stress sensors. All endogenous proteins were prepared from wild-type non-stressed CHO cells. P1, nuclear; Mito, mitochondrial; P3, microsomal; Cyt, cytosolic fractions. NucleoP, nucleoporin p62; Complex V, complex V ATP synthase inhibitor; Cyto c, cytochrome c; ERp57, ER thiol-disulfide oxidoreductase p57; ERK, extracellular signal-regulated kinase. Phosphatidylserine (PtSer) synthase activity was measured by the autoradiographic measurement of 14C-PtSer as described [10]. All other proteins were measured by immunoblotting. (b) The subcellular distribution of IRE1 in CHO cells with reduced expression of mitofusin-2 (MFN2) or Sig-1Rs. Control (siCon) or specific siRNAs (siMFN2, siSig-1R) were transfected two days before the membrane fractionation. (c) Confocal microscopic observation of the subcellular distribution of Sig-1Rs and IRE1 in CHO cells. In top panels, endogenous ERp57 and transfected full-length IRE1-V5 were immunostained. Asterisks indicate CHO cells transfected with IRE1-V5 (Note: no V5 immunoreactivity in non-transfected cells, verifying the high selectivity of V5 immunostaining). In bottom panels, GFP targeting mitochondria was expressed by gene transfection. Arrows indicate clusters of IRE1-FLAG apposing mitochondria. Scale = 10 µm. Insets on a 5× magnification.
Figure 2.
Sigma-1 receptors stabilize IRE1.
(a) Sig-1R knockdown decreases phosphorylated IRE1 (pIRE1) in various types of cells when cells are under ER stress. Two days after the transfection of control siRNA (siCon) or Sig-1R siRNA (siSig-1R), cells were treated with thapsigargin (Tg) at 1 µM for 60 min. IRE1 were immunoprecipitated from 60–1000 µg of total protein lysates. (b) The temporal course of pIRE1 levels during ER stress. Control or Sig-1R siRNA was transfected to CHO cells two days before Tg. pIRE1 was measured by immunoprecipitation. The level of pIRE1 (partially phosphorylated IRE1 plus hyperphosphorylated IRE1) was normalized to ERK. The graph represents the means±S.E.M. ***p<0.001 (n = 6). (c) Effect of the Sig-1R knockdown on the level of pIRE1 in CHO cells either overexpressing LDL receptor G544V mutants or being under the heatshock treatment. LDL receptors G544V (LDLR G544V) were induced by the tetracycline treatment for 16 hrs (TC; see Methods). The level of pIRE1 was measured by immunoprecipitation and was normalized to ERK. Graphs represent means±S.E.M (n = 4). *p<0.05, **p<0.01. (d) Effects of lactacystin (10 µM; applied 10 min before Tg) or kifunensin (2 µg/ml; applied 10 min before Tg; lower panels) on Tg (1 µM for 1 hr)-induced decrease of pIRE1 in Sig-1R knockdown CHO cells. (e) Pulse-chase experiment measuring the effect of Sig-1R knockdown on the life-time of IRE1 when cells were under ER stress. After labeling with 35S-methionine, CHO cells transfected with control or Sig-1R siRNA were chased in the presence of Tg (1 µM) for indicated periods of time. IRE1 was immunoprecipitated and then detected by direct autoradiography. The graph represents mean±S.E.M (n = 4). **p<0.01 compared with siCon at the same time point.
Figure 3.
Sigma-1receptors preferentially associate with the monomeric form of IRE1 in the lumen of the ER.
(a) Time-dependent association of V5-tagged Sig-1Rs with IRE1 during ER stress. IRE1 was immunoprecipitated in CHO cells expressing V5-tagged full-length Sig1R-V5 (1–223) or truncated Sig-1R-V5 (1–50). Thapsigargin (Tg) was applied to the medium for indicated periods of time. (b) Direct association between purified ΔIRE1-V5/His and the ER lumenal domain (116–223) of Sig-1Rs (GST-Sig-1R116–223) in vitro. Purified ΔIRE1 immobilized on a Ni+-column was incubated with purified GST-Sig-1R116–223 polypeptides. After extensive washing, the GST-Sig-1R116–223 associating with ΔIRE1-V5/His was measured by Western blotting. β-mer (+): the purified ΔIRE1-V5/His was pretreated with β-mercaptoethanol to prevent the dimerization of ΔIRE1-V5/His. The image represents three independent experiments. (c) Involvement of both the ionic bond (D123 site) and the disulfide bonds in the dimerization of IRE1. To assess the dimerization of IRE1, FLAG-tagged full-length IRE1 was co-immunoprecipitated with ΔIRE1-V5 with or without a point-mutation at D123 (e.g., ΔIRE1, D123PΔIRE1). Dimerization of IRE1, via disulfide bonds, was disrupted by DTT (20 mM for 30 min). (d) Association of Sig-1Rs with IRE1 depends on the dimerization status of IRE1. V5-tagged ΔIRE1 with or without a point-mutation at D123 were immunoprecipitated. The association of Sig-1R with Δ IRE1-V5 was assessed by measuring co-immunoprecipitated Sig-1R-FLAG. The graph represents means±S.E.M. *p<0.05 (n = 4). (e) Effects of overexpressed Sig-1Rs on the dimerization of IRE1. Dimerization of IRE1 was assessed by measuring IRE1-FLAG co-immunoprecipitated with V5-tagged IRE1 (ΔIRE1-V5 or D123PΔIRE1-V5). Note: overexpressed Sig-1Rs co-immunoprecipitated significantly more with IRE1-V5 when IRE1 was mutated at D123. The graph represents means±S.E.M (n = 4, *p<0.05).
Figure 4.
Effect of Sig-1R knockdown on the Tg-stimulated IRE1/XBP1 signaling pathway in CHO cells.
(a) Effect of Sig-1R knockdown on Tg-induced apoptosis. Induction of apoptosis by Tg (1 µM, for 24 hr) was quantified by using Hoechst 33342 staining. The percentage of apoptotic cells was measured from 6 individual samples (more than 200 cells were counted in each sample) and is reported as means±S.E.M. ***P<0.001 compared with vehicle (Veh), ##P<0.01 compared with siCont with Tg. (b) Effects of Sig-1R knockdown on the splicing of XBP1 mRNA. CHO cells transfected with control or Sig-1R siRNA were treated with Tg for indicated periods of time. Total RNA was extracted and XBP-1 or actin transcripts were amplified by RT-PCR. Un, unspliced; S, spliced; sXBP1; spliced XBP1. The graph represents means±S.E.M. **p<0.01 compared with siCon at the same time point (n = 4). (c) Effects of Sig-1R siRNA on the expression of FLAG-XBP1-venus fusion proteins. CHO cells transfected with FLAG-XBP1-venus plasmids were treated with Tg, and the expression of FLAG-XBP1-venus was measured by Western blotting (left panels) or a fluorescence microplate reader (the right graph). *p<0.05, ***p<0.001 compared with siCon at the same time point (n = 8). Note that venus is expressed only when FLAG-XBP1 mRNA is spliced by active IRE1. (d) Sig-1R siRNA enhanced apoptosis in Tg-treated cells negatively correlates with the activity of XBP1 mRNA splicing. Fluorescence intensities (annexin V vs. venus) of individual CHO cells were plotted in the graph. Control siRNA (siCon) in green, Sig-1R siRNA (siSig-1R) in orange. (e) Overexpression of spliced XBP1 attenuates apoptosis enhanced by Sig-1R knockdown in Tg-treated cells. Each bar represents the means±S.E.M. (n = 6–7 samples; >50 cells/sample were counted). *P<0.05, ***P<0.001 compared with siCon alone. ##P<0.01 compared with siSig-1R with Tg.
Figure 5.
Mitochondria-derived oxidative factor activates IRE1.
(a) Effects of Tg, antimycin A (AMA), or rotenone on the activation of ER stress sensors. IRE1 were immunoprecipitated before detection by western blottings. No active form of ATF6 (p50) or phosphorylated PERK was detected under the experimental condition with AMA or rotenone. (b) Effects of NAC on phosphorylation of IRE1 caused by AMA. (c) Effects of AMA on the splicing of the XBP1 mRNA in CHO cells. Un, unspliced; S, spliced. (d) AMA (1 µM) induced the expression of XBP1-venus in an IRE1-dependent manner. CHO cells were transfected with FLAG-XBP1-venus plasmids and siRNA (siCon, scrambled siRNA; siIRE1, IRE1 siRNA) two days before the AMA treatment. Induction of FLAG-XBP1-venus proteins was measured by a fluorescence microplate reader. ***P<0.001 (n = 12), compared with siCon. (e) Selective activation of IRE1 by AMA in the MAM-containing crude mitochondrial fraction. AMA-treated CHO cells were homogenized and subjected to differential centrifugation. The asterisk indicates the hyperphosphorylated form of IRE1. (f) Effect of mifotusin-2 knockdown on the AMA-induced IRE1 phosphorylation. siRNA against mitofusin-2 (siMFN2) or scrambled control siRNA (siCon) was transfected into CHO cells two days before the treatment with AMA. In the graph, IRE1 was normalized to ERK and shown as means±S.E.M. **p<0.01 (n = 4) compared with siCon without AMA. (g) AMA induces a reduction of pIRE1 proteins in CHO cells with reduced expression of Sig-1Rs. Control or Sig-1R siRNA were transfected into CHO cells two days before AMA. IRE1 was immunoprecipitated before detection.
Figure 6.
Schematic model depicting the role of the Sig-1R chaperone in the activation of IRE1.
The Sig-1R molecular chaperone enhances its association with IRE1 to correct or stabilize the conformation of IRE1 when cells are facing ER stress (i.e., as indicated in the blue-lined rectangle). This transient association of the Sig-1R with IRE1 interferes with the dimerization of IRE1, leading to a delay in the transautophosphorylation of IRE1 (30 min). This delayed dimerization/phosphorylation, however, ensures a long-lasting active form of IRE1 (the cytoplasmic domain filled in red) which splices the XBP1 mRNA. In lower panels, when Sig-1R knockdown cells encounter ER stress, IRE1, although being misfolded, can still quickly dimerize and transautophosphorylate (5–15 min lower panels). The conformationally awry pIRE1, which may still possess endonuclease activity albeit being less compared to controls, is however readily ubiquitinated and degraded by proteasomes.