The ALS-linked E102Q mutation in Sigma receptor-1 leads to ER stress-mediated defects in protein homeostasis and dysregulation of RNA-binding proteins

Abstract

Amyotrophic lateral sclerosis (ALS) is characterized by the selective degeneration of motor neurons (MNs) and their target muscles. Misfolded proteins which often form intracellular aggregates are a pathological hallmark of ALS. Disruption of the functional interplay between protein degradation (ubiquitin proteasome system and autophagy) and RNA-binding protein homeostasis has recently been suggested as an integrated model that merges several ALS-associated proteins into a common pathophysiological pathway. The E102Q mutation in one such candidate gene, the endoplasmic reticulum (ER) chaperone Sigma receptor-1 (SigR1), has been reported to cause juvenile ALS. Although loss of SigR1 protein contributes to neurodegeneration in several ways, the molecular mechanisms underlying E102Q-SigR1-mediated neurodegeneration are still unclear. In the present study, we showed that the E102Q-SigR1 protein rapidly aggregates and accumulates in the ER and associated compartments in transfected cells, leading to structural alterations of the ER, nuclear envelope and mitochondria and to subsequent defects in proteasomal degradation and calcium homeostasis. ER defects and proteotoxic stress generated by E102Q-SigR1 aggregates further induce autophagy impairment, accumulation of stress granules and cytoplasmic aggregation of the ALS-linked RNA-binding proteins (RBPs) matrin-3, FUS, and TDP-43. Similar ultrastructural abnormalities as well as altered protein degradation and misregulated RBP homeostasis were observed in primary lymphoblastoid cells (PLCs) derived from E102Q-SigR1 fALS patients. Consistent with these findings, lumbar α-MNs of both sALS as well as fALS patients showed cytoplasmic matrin-3 aggregates which were not co-localized with pTDP-43 aggregates. Taken together, our results support the notion that E102Q-SigR1-mediated ALS pathogenesis comprises a synergistic mechanism of both toxic gain and loss of function involving a vicious circle of altered ER function, impaired protein homeostasis and defective RBPs.

Figure 1

mSigR1 abnormally accumulates in the ER and induces cellular toxicity. (a) Model of the homotrimeric SigR1 (based on ref. 16). Subunits are represented in gray, secondary structure of one subunit is color-coded; each monomer is composed of an amino-terminal transmembrane domain (red) which crosses the ER membrane from lumen to cytosol. The transmembrane domain is followed by two α-helices (green) that lead to a β-barrel (light violet), the putative ligand binding domain. Two carboxy-terminal α-helices (blue) form a hydrophobic membrane-embedded surface. The location of the critical residue Glu-102 (E102Q) is marked by a red dot. Red star and red triangle represent two recently discovered mutations. (b) Reticular pattern of SigR1 staining combined with a distinct nuclear envelope localization (lower panel) in wtSigR1-transfected MCF-7 cells; globular mSigR1 aggregates (arrows; see also enlargement, middle row) in MCF-7 cells expressing mSigR1. Immunofluorescence; scale bars, 15 μm. (c) Co-immunofluorescence of wtSigR1 and mSigR1 with emerin as a nuclear envelope marker in MCF-7 cells (arrowheads). Note that the focal emerin accumulations (arrows) co-localize with SigR1 aggregates. Scale bar, 10 μm. (d and e) Co-localization of wtSigR1 and mSigR1 with the ER markers KDEL and (e) calreticulin in MCF-7 cells. Scale bar, 15 μm. (f and g) Co-labelling of SigR1 with the β-cop (ER-Golgi-associated compartments) and GM130 (Golgi marker). Note the co-localization of mSigR1 with β-cop and the Golgi dispersal (arrowheads) in mSigR1 expressing MCF-7 cells. Scale bar, 10 μm. (h) Co-immunofluorescence of wtSigR1 and mSigR1 with mito-red as a marker for mitochondria in MCF-7 cells. Scale bar, 10 μm. (i) Quantification of the mito-red staining depicted in (h) showing numbers of enlarged mitochondria/random field of view. (j) NSC34 and MCF-7 cells were co-transfected either with empty pcDNA vector, wtSigR1 or mSigR1 together with the luciferase gene downstream of the ERSE promoter. After 48 h, luciferase units (AU) were analyzed as a measure of ER stress. *P<0.05,**P<0.005, #not significant. (k) MCF-7 cells were transfected with pcDNA, wtSigR1 or mSigR1 as described above and analyzed for ER stress and UPR induction by immunoblotting. (l) Quantification of the band intensities normalized with α-tubulin depicted in k. Values represent the mean±S.D. of three independent experiments. *P<0.05. (m) Ubiquitin immunoreactivity of wtSigR1 and mSigR1 in MCF-7 cells. Scale bar, 10 μm. (n) Co-localization of mSigR1 aggregates with the accumulated 20s subunit of the proteasome in MCF-7 cells expressing mSigR1. Scale bar, 10 μm. (o) Chymotrypsin-like proteasomal activity assays were performed using MCF-7 cells transfected with pcDNA, wtSigR1 and mSigR1, as described in material and methods. Note the significant decrease in proteasome activity due to mSigR1 expression when compared to wtSigR1 and pcDNA. Means±S.D. of three independent experiments each performed in triplicate. *P<0.05, **P<0.005, # not significant

Figure 2

mSigR1 is abnormally accumulated in the ER and induces cellular toxicity in E102Q-SigR1 fALS patient lymphoblastoid cells. (a) Immunoreactivity of globular SigR1 aggregates (arrows) in E102Q-SigR1 fALS patient lymphoblastoid cells compared to the healthy control. Note the co-localization of SigR1 aggregates with the nuclear envelope marker emerin (arrowhead). Scale bar, 15 μm. (b and c) Co-localization of mSigR1 aggregates with the ER markers KDEL and (c) calreticulin in E102Q-SigR1 fALS patient lymphoblastoid cells compared to the healthy control. Scale bars, 15 μm. (d and e) Upregulation and co-localization of the ER stress markers GRP78 (d) and pPERK (e) with mSigR1 aggregates in E102Q-SigR1 fALS patient lymphoblastoid cells compared to the healthy control. Scale bars, 15 μm. (f) Immunoblot analysis of SigR1 and established ER stress and UPR markers in healthy control and E102Q-SigR1 fALS lymphoblastoid cell lysates. Dot blot analysis (top) shows the presence of triton-X insoluble aggregates in E102Q-SigR1 fALS lymphoblastoid cells. (g) Quantification of the band intensities normalized with α-tubulin depicted in f. Values represent the mean±S.D. of three independent experiments. *P<0.05. (h–i) RT-PCR analysis of the UPR pathways in three healthy control lymphoblastoid cell lines compared to two E102Q-SigR1 fALS patient lymphoblastoid cell lines. E102Q-SigR1 fALS patient’s lymphoblastoid cells showed a significant increase in ATF4 mRNA expression. *P<0.05, # not significant. Note that SigR1 gene expression does not differ between controls and E102Q-SigR1 fALS patients. (j) E102Q-SigR1 fALS patient’s lymphoblastoid cell (n=2) and control lymphoblastoid cell (n=3) lysates were subjected to chymotrypsin-like proteasomal activity and caspase-3 activity assays as described in material and methods. Values are derived from the average of three control lymphoblastoid cell lines compared to average of two E102Q-SigR1 fALS patient’s lymphoblastoid cell lines from three independent experiment,**P<0.005. (k) GM130 and SigR1 immunolabelling in E102Q-SigR1 fALS and control lymphoblastoid cells. Scale bar, 15 μm. (l) Immunoblot analysis using SigR1 antibody to compare SigR1 levels in lymphoblastoid cells from E102Q-SigR1 patient lymphoblasts and healthy controls as well as transiently transfected MCF-7 cells. Quantification of the band intensities normalized with α-tubulin is shown below. (#) denotes absence of significant differences

Figure 3

mSigR1 induces mitochondrial toxicity and fails to mobilize IP3R and SOCE-mediated Ca²⁺ signalling. (a and b) MCF-7 cells were transiently transfected as previously described; 48 h later cells were loaded with Fura-2AM for 30 min, washed twice and then stimulated with 10 μM BDK under Ca²⁺-free conditions. Average traces of the BDK-induced increase in [Ca²⁺]ⁱ from the ER store through IP3R are represented. Note the increase of [Ca²⁺]ⁱ in wtSigR1-transfected cells (green curve) compared to a significant decrease in mSigR1-transfected cells (red curve). Results are expressed as mean±S.E.M. of ~30 cells. Mean changes in peak [Ca²⁺]ⁱ measured are given. The asterisks denote a statistically significant difference (*P<0.05). (c) Representative SOCE elicited in transfected cells loaded with Fura-2AM as described above. SOCE was triggered by the addition of 2 μM ionomycin under Ca²⁺-free conditions. SOCE developed (vertical dot line) after the addition of Ca²⁺ is depicted. (d) STIM1 immunoblot analysis from lysates obtained from pcDNA, wtSigR1 and mSigR1-transfected MCF-7 cells. Note the significant down-regulation of STIM1 in mSigR1-transfected cells. The fold change below represents the quantification of band intensities normalized against α-tubulin. Values derived from three independent experiments. *P<0.05. (e) Significantly decreased STIM1 levels in E102Q-SigR1 fALS lymphoblastoid cell lysates compared to healthy control lymphoblastoid cells. The fold change below represents the quantification of band intensities normalized against α-tubulin. Values derived from three independent experiments. *P<0.05. (f) Significantly reduced mitochondrial membrane integrity and ATP production in mSigR1 expressing MCF-7 cells compared to wtSigR1 expressing cells measured by the tox glow assay. Values derived from three independent experiments. *P<0.05. (g) JC-1 staining of HeLa cells transfected with wtSigR1 or mSigR1. Note the reduced mitochondrial potential in mSigR1 expressing cells. Scale bar, 10 μm. (h) JC-1 staining of lymphoblastoid cells obtained from E102Q-SigR1 fALS patients and healthy controls. Note the decreased membrane potential in mSigR1 expressing cells. Scale bar, 10 μm. (i) Cytochrome C immunolabelling of HeLa and MCF-7 expressing wtSigR1 or mSigR1. Note the Cytochrome C release in cells showing mSigR1 aggregates (arrowheads). Scale bar, 10 μm

Figure 4

mSigR1 leads to structural abnormalities of ER and mitochondria. (a) MCF7 cells expressing pcDNA, wtSigR1 or mSigR1 were fixed with 2.5% buffered glutaraldehyde and processed for EM. Several membrane-bound vacuolar structures (black arrows) probably derived from the ER in a representative mSigR1-expressing cell compared to cells expressing wtSigR1 and pcDNA control. Scale bars, 1 μm. (b) Higher magnification showing the widened ER (arrows in right panel) in representative mSigR1-expressing cells compared to pcDNA and wtSigR1-transfected cells. Scale bar, 0.5 μm. (c) Higher magnification showing double membrane autophagic vacuoles in representative mSigR1-expressing MCF-7 cells. (a) Large autophagosome (AV; arrowheads) containing mitochondria and other structures in close proximity to lysosomes (Lys); (b) autophagosome (arrowhead) and widened ER (white arrow); (c) autophagosome (AV) and lysosomes (Lys). Scale bar, 0.5 μm. (d) Enlarged mitochondria (arrows) showing abnormal cristae architecture, some undergoing mitophagy (arrowheads) in mSigR1-transfected cells. Scale bar, 0.4 μm. (e) Primary lymphoblastoid cells from healthy control and E102Q-SigR1 fALS patients were fixed with 2.5% buffered glutaraldehyde and processed for EM. Control lymphoblasts show an overall normal ultrastructure, whereas E102Q-SigR1 fALS patient lymphoblastoid cells reveal prominent accumulation of autophagic material. Scale bars, 2.5 μm. (f) Prominent nuclear envelope (arrows) protrusions in E102Q-SigR1 fALS lymphoblastoid cells and rather normal nuclear envelope in control lymphoblastoid cells. Scale bar, 0.5 μm. (g) Normal ER (white arrows) and mitochondria (gray arrows) in healthy control lymphoblastoid cells; E102Q-SigR1 fALS patient lymphoblastoid cells showing overall widened ER (arrows). Scale bar, 0.5 μm. (h) E102Q-SigR1 fALS lymphoblasts displaying vacuolar degeneration of mitochondria (gray arrows), multivesicular bodies (arrowheads) and autophagosomes filled with membranous and granular material (white arrows). Black arrow: nuclear envelope protrusion. Scale bars, 0.4 μm

Figure 5

mSigR1 leads to defective autophagy and autophagosome-lysosome fusion. (a) Immunoblot analysis for autophagy markers in A431 cells transiently transfected with pcDNA, wtSigR1 and mSigR1. Note the increased levels of autophagy markers in mSigR1 expressing cells. (b–d) MCF-7 cells were transiently transfected with wtSigR1 or mSigR1 and then processed for (b) co-immunolabelling using SigR1 and LC3 antibodies, (c) co-immunolabelling using SigR1 and Cyto-ID dye (for labelling autophagosomes) and (d) co-immunolabelling using SigR1 and p62 antibodies. Scale, 10 μm. (e) Increased accumulation of autophagosomes in the stable autophagy reporter cell line NIH3T3-GFP transfected with mSigR1. Green: GFP-LC3; red: SigR1; scale bar, 10 μm. (f) Immunoblot analysis of NIH3T3-GFP-LC3 cells transfected with wtSigR1 or mSigR1 and additionally treated with the autophagy inhibitor Bafilomycin A for 2 h. Note the unchanged LC3-II levels in mSigR1-transfected cells after Bafilomycin A treatment. Corresponding densitometric data are shown at the bottom; where the upper number represents the relative LC3II/Tub levels and the lower numbers are the S.D. The asterisks (*) denote significant differences (*P<0.05), while # denotes absence of a significant difference. (g and h) Primary fibroblasts isolated from autophagy reporter GFP-LC3 transgenic mice were transfected with wtSigR1 or mSigR1 and immuno-labelled by SigR1 (g) and p62 (h) antibody (merge image for GFP-LC3, green, and SigR1/p62, red). Note the localization of SigR1 at the periphery of this particular cell (g, arrowheads) and the co-localization of GFP-LC3 with globular p62 accumulations. Scale bar, 10 μm. (i) Co-localization of SigR1 with p62 and LC3 in E102Q-SigR1 fALS lymphoblastoid cells compared to healthy control lymphoblastoid cells. (j) Cyto-ID (green) staining (right) showing the accumulation of autophagosomes in E102Q-SigR1 fALS lymphoblastoid cells. (k) Immunoblot analysis of established autophagy markers in E102Q-SigR1 fALS lymphoblastoid cells in comparison to healthy controls. (l) A431 cells were transfected as described above. Forty-eight  hours later, transfected cells were processed for the EGFR degradation assay as described in Materials and Methods and analyzed by immunoblotting with the EGFR antibody. Note the delayed EGFR degradation in mSigR1 expressing cells. (lower) Quantification of immunoblot analysis. Values are expressed as mean±S.D. from three independent experiments. *P<0.05. (m) NIH3T3 cells expressing RFP-GFP-LC3 were transfected with pcDNA, wtSigR1 or mSigR1. Forty-eight hours later the fusion of autophagosomes with lysosomes was measured by live cell imaging. Scale bar, 25 μm. (lower) The rate of autophagosome maturation reflected by the Pearson coefficient (green/red fluorescence ratio) at each time point indicated. Values are represented as means±S.E.M. of triplicate experiments *P<<0.0001. (n) EM picture of NIH-3T3 cells stably expressing RFP-GFP-LC3 transfected with mSigR1. Note the autophagosome (white arrow) and lysosomes (gray arrow) without any fusion

Figure 6

mSigR1 impairs ER to Golgi transport. (a) Cos7 cells showing abnormal SigR1 accumulations (arrows) after mSigR1 transfection and normal ER localization of SigR1 (right) after wtSigR1 transfection. Scale bar, 15 μm. (b) Cos7 cells were co-transfected with VSVG-GFP together with pcDNA, wtSigR1 or mSigR1. 48 h after transfection, fluorescence associated with the Golgi complex was photobleached (FRAP, see Materials and Methods) with a high-intensity laser beam. Subsequently, the inward delivery of VSVG-GFP from pre-Golgi intermediates was monitored for the indicated periods of time. Scale bar, 10 μm. (c) Fluorescence recoveries after photobleaching curves of mSigR1-transfected cells show a clear decrease of the mobile fraction as compared to cell transfected with pcDNA or wtSigR1. Error bars indicate the S.E.M.; (right) the comparison of the mobile fractions in pcDNA, wtSigR1 and mSigR1 expressing cells. In the box plots, the line in the middle of the box indicates the median; the top line indicates the 75th quartile, whereas the bottom line indicates the 25th quartile. Whiskers represent the 10th and 90th (upper) percentile, respectively. (d) mSigR1 aggregates (arrowheads) co-localize with the endosomal markers Rab5 (upper) and Rab7 (lower) in Cos-7 cells transfected with wtSigR1 and mSigR1, respectively. Scale bar, 10 μm. (e) Co-localization of mSigR1 aggregates with the endosomal markers EEA1 and Rab7 in E102Q-SigR1 fALS patients' lymphoblastoid cells compared to healthy controls. Scale bars, 15 μm. (f) Immunoblot analysis of Cos-7 cells transfected with pcDNA, wtSigR1 and mSigR1. (g) Quantification of band intensities normalized against α-tubulin depicted in (f). Values are expressed as mean±S.D. from three independent experiments. *P<0.05

Figure 7

mSigR1 accumulation leads to altered RNA-binding protein homeostasis. (a) Co-staining of SigR1 and TDP-43 in MCF-7 cells transfected with either wtSigR1 or mSigR1. Note the minor cytoplasmic TDP-43 accumulations (white arrow) without co-localization with SigR1, in contrast to the more pronounced TDP-43 aggregates co-localizing with SigR1 in E102Q-SigR1 fALS lymphoblastoid cell depicted in e. Scale bar, 15 μm. (b) Translocation of FUS from the nucleus (yellow arrow) to the cytoplasm (white arrow) and co-localization with SigR1 aggregates in MCF-7 cells expressing mSigR1 compared to wtSigR1-transfected cells. Scale bar, 10 μm. (c) Immunofluorescence staining of SigR1 and matrin-3 in MCF-7 cells expressing wtSigR1 or mSigR1. Note the cytoplasmic accumulations (white arrow) along with the loss of nuclear matrin-3 (yellow arrow) corresponding to the higher amount of globular mSigR1 aggregates. Scale bar, 10 μm. (d) Quantification (n=3 each) of the experiments illustrated in (a–c). *P<0.05. (e–g) Nuclear loss of TDP43 (e), FUS (f) and matrin-3 (g) and their co-localization with mSigR1 aggregates in E102Q-SigR1 fALS lymphoblastoid cells compared to healthy controls. Scale bar, 10 μm. (h) Immunoblot analysis of subcellular fractions obtained from E102Q-SigR1 patients' lymphoblastoid cell lines compared to healthy controls (n=2 for fALS; n=3 for control). Matrin-3, FUS and TDP-43 distributions are shown in the cytoplasmic (Cyto.), membrane (Memb.) and nuclear (Nucl.) fractions. Note the translocation of matrin-3 from the nucleus to the cytoplasm in E102Q-SigR1 patients' lymphoblastoid cells. Tubulin is used as a loading control and Lamin A as a positive control for the nuclear fraction. (i) Immunohistochemical analysis of lumbar α-MNs using matrin-3 antibody in sALS and fALS compared to normal controls. Cytoplasmic accumulation (arrowheads, middle) of matrin-3 in α-MNs of fALS patients harbouring FUS and C9ORF72 mutations. Strong nuclear immunoreactivity of matrin-3 (arrows) is evident in sALS α-MNs (right), whereas the control shows an overall weaker nuclear matrin-3 signal (arrows, left). Paraffin sections, DAB-immunohistochemistry; scale bars, 20 μm. (j) Quantification of immunohistochemical data illustrated in I; for the immunohistochemical analysis n=12 sALS; n=8 (C9ORF72); n=4 (FUS) and n=4 control cases were examined. *P<0.05, while # denotes absence of significance (k) loss of nuclear Tia1 (yellow arrows) and accumulation of cytoplasmic Tia-1 (white arrow in the middle panel) positive stress granules co-localized with mSigR1 aggregates in MCF-7 cells. Scale bar, 15 μm. (l) Quantification of data shown in (k). *P<0.05

Figure 8

Illustration showing the proposed mechanisms of ALS pathogenesis associated with the E102Q mutation in SigR1