Introduction
Parkinson’s disease (PD), the second-most common neurodegenerative disease, is characterized by selective loss of dopaminergic (DA) neurons of the substantia nigra pars compacta (SNpc), accumulation of intracellular inclusions containing α-synuclein, and subsequent progressive impairment of dopaminergic neurons—the clinical feature of PD [
6,
58]. It is mostly sporadic; less than 10% of PD cases are inherited [
28]. Despite intensive investigation, current treatments, medications, and even surgery do not cure or stop the progression of PD, highlighting the critical importance of understanding the mechanisms involved in the loss of dopaminergic neurons in PD. Recent studies suggested that non-neuronal cells, such as astrocytes, accumulate α-synuclein during PD, and can contribute to neurodegeneration through various pathways [
7,
56], suggesting astrocytes play critical roles in neuronal dysfunction and interplays between astrocytes and neurons may provide insights into neuronal dysfunction and death in PD. However, relationship between astrocytic α-synucleintoxicity and neuronal dysfunction remains unclear.
Accumulating evidence from studies of the human PD brain suggest that the progressive deterioration of vulnerable SNpc DA neurons arises from cellular disturbances produced by misfolding and aggregation of the synaptic protein α-synuclein, disruption of the autophagy-lysosome system, mitochondrial dysfunction, and/or endoplasmic reticulum (ER) stress [
31]. The ER is the subcellular site of protein folding and maturation, and the main intracellular Ca
2+ store of the cell. Since ER resident chaperones are involved in protein folding require high Ca
2+ concentrations for their activity, altered ER Ca
2+ homeostasis can result in an imbalance between the capacity of the protein processing machinery and the amount of unfolded proteins requiring processing, leading to an accumulation of unfolded proteins and ER stress [
24,
30].
The increasing number of unfolded proteins inside the ER lumen provokes the dissociation of grp78/BiP (78 kDa glucose-regulated protein) from three ER transmembrane receptors—PKR-like endoplasmic reticulum kinase (PERK), activating transcription factor 6 (ATF6) and inositol-requiring enzyme 1 (IRE1)—thereby initiating the unfolded protein response (UPR). This ER-specific stress response serves to maintain cell survival through different molecular pathways, including reduced translation-initiation rate, enhanced protein folding, and/or elimination of misfolded proteins [
53]. Under conditions of prolonged or severe ER stress, however, the UPR switches from homeostatic feedback regulation towards proapoptotic signaling [
52]. Growing evidence from recent studies indicates that the accumulation of misfolded proteins in the brain is a salient feature of most neurodegenerative diseases, including Alzheimer’s disease, amyotrophic lateral sclerosis, Huntington’s disease, and PD [
29]. These diseases are now classified as protein misfolding disorders (PMDs) [
13,
57]. The mechanisms leading to ER stress in PMDs and the actual impact of the UPR on the degeneration cascade in this disease are just starting to be uncovered.
A major genetic form of PD is caused by mutations in leucine-rich repeat kinase 2 protein (LRRK2) [
37,
66]. The G2019S (GS) mutation within the kinase domain encoded by exon 41 is the most common mutation of LRRK2, which alters LRRK2 GTPase and kinase activities and accounts for ~ 1% of sporadic PD and up to 25% of familial PD in certain populations [
28]. The LRRK2 G2019S mutation causes a gain of function effect that could involve alterations in autophagy-lysosomal [
61] and microRNA [
10] pathways, and dysregulation of protein quality control [
18,
32], oxidative stress [
32,
49], and protein synthesis [
16] mechanisms. Although the function of LRRK2 in ER stress remains a matter of debate, studies in
Caenorhabditis elegans have demonstrated that expression of wild-type LRKK2 protects dopaminergic neurons against neurotoxicity induced by human α-synuclein through upregulation of grp78/BiP [
65]. Another study using a
C. elegans model lacking the LRRK2 homolog suggested that LRRK2 is critical for preventing ER stress and spontaneous neurodegeneration [
50]. It has also been suggested that LRRK2 regulates anterograde ER-Golgi transport by anchoring Sec16A at ER exit sites, leading to a reduction in ER stress [
3]. Despite these interesting findings, the possible contribution of ER stress to the pathogenic manifestations of mutant LRKK2 in mammalian cells has not yet been addressed.
In this study, we showed that the LRRK2 G2019S mutant is responsible for ER stress in α-synuclein–treated brain astrocytes. Immunostaining and subcellular fractionation revealed that LRRK2-G2019S dissociates from 14 to 3-3 s and then localizes to the ER membrane. Using mass spectrometry (MS) proteomic screening of LRRK2-associated proteins, we identified sarco/endoplasmic reticulum Ca2+-ATPase (SERCA) as a protein that strongly interacts with the LRRK2 G2019S mutant in the ER. Binding of LRRK2 G2019S to SERCA inactivated SERCA by maintaining the interaction of SERCA with phospholamban (PLN), which negatively regulates SERCA activity through direct association. The inactivation of SERCA by LRRK2 G2019S led to ER Ca2+ depletion, followed by chronic ER stress, mitochondria dysfunction, and cell death. Collectively, these findings indicate that LRRK2-G2019S accelerates ER stress, suggesting a molecular basis for the pathogenesis of PD in patients harboring this mutant.
Materials and methods
Animals
G2019S-LRRK2-Tg FVB mice were purchased from Jackson Laboratory (stock #009609, Bar Harbor, ME, USA). Non-Tg and G2019S-LRRK2-Tg heterozygous mice were prepared by crossing G2019S-LRRK2 heterozygous mice with wild-type mice. Genotyping was carried according to the vendor’s instructions. All animal procedures were approved by the Ajou University Institutional Animal Experimentation Committee (AMC-119).
Cell culture
Primary astrocytes were cultured from the cerebral cortices of 1-d-old non-Tg and G2019S-LRRK2-Tg heterozygous mice. Briefly, cortices were triturated into single cells in Dulbecco’s modified Eagle’s medium (DMEM; Sigma-Aldrich, St. Louis, MO, USA) containing 10% (v/v) fetal bovine serum (FBS; Hyclone, South Logan, UT, USA), plated into 75 cm2 T-flasks, and incubated for 2 wk. Following removal of microglia, primary astrocytes were enzymatically dissociated with trypsin (Sigma) for 5 min at 37 °C in a humidified 5% CO2, 95% air chamber. Trypsinization was quenched by adding astrocyte culture medium and centrifuged (~ 200 g) for 5 min. Microglia and meningeal cells were depleted by incubating astrocytes with serum-free DMEM for 2 d before use. The cell populations obtained consisted of more than 95% authentic astrocytes, as determined using the astrocyte marker GFAP (glial fibrillary acidic protein) immunofluorescence.
Primary neurons were cultured from embryonic mouse cortices (E17). Briefly, cortices were dissected in Hank’s Buffered Salt Solution (HBSS; Gibco, Carlsbad, CA, USA) supplemented with HEPES (10 mM, pH 7.4). Tissues were incubated in HBSS containing trypsin (Gibco) and DNase I (100 μg/ml) for 15 min at 37 °C, then dissociated by gentle pipetting. Dissociated cells were plated in poly-D-lysine (1 mg/ml)-coated 6-well plate (3 × 105 cells/well) or 12-mm cover glasses (1 × 104 cells) in Neurobasal medium containing B27 (2%), sodium pyruvate (1%), penicillin/streptomycin (1%), and GlutaMax (1%) (all supplements were from Gibco). Cells were incubated for 10 d and then challenged with α-synuclein for 24–48 h. For co-culture with astrocytes, cortical neurons were seeded on non-Tg or LRRK2-GS astrocyte monolayers for 10 d, then treated with α-synuclein for 24–48 h and assayed by immunocytochemistry.
C2 myoblasts were cultured as previously described [
22]. Briefly, cells were cultured in DMEM supplemented with 10% (
v/v) FBS. Myogenic differentiation was induced by incubating in low-serum medium (DMEM plus 1% FBS) for 3 d. To induce ER stress, the astrocytes, neuron, and C2 myoblasts were incubated with vehicle or tunicamycin (0.1 μg/mL) (Sigma) for 3 ~ 24 h for each separate experiment.
Organotypic brain slice culture
Organotypic brain slice cultures were prepared from postnatal day 7–8 non-Tg and LRRK2-GS mice. Mice were anesthetized, and cortices were dissected and coronally sectioned (400 μm) using a McIlwain tissue chopper (Mickle Laboratory Engineering, UK). Slices were mounted on Millicell cell culture inserts (0.4 μm pore size, 30 mm diameter; Millipore, Burlington, MA, USA). Culture medium (50% minimal essential medium [MEM] containing 25% HBSS, 25% heat-inactivated horse serum, 0.5% glucose, 1 mM l-glutamine) was changed every 2–3 d. Slices were treated with α-synuclein or tunicamycin after 7 d in culture.
α-Synuclein
The human recombinant α-synuclein was purchased from rPeptide (Watkinsville, GA, USA) or was a kind gift from Professor Sang Myun Park (Ajou University, Korea). Purified recombinant α-synuclein protein was stored at − 80 °C until use as monomeric α-synuclein. The α-synuclein oligomer was prepared as previously described [
4] with minor modifications. Briefly, 1 mg/ml of monomeric α-synuclein was dissolved in PBS (0.01 M sodium phosphate (pH 7.4), 150 mM NaCl), and incubated at 37 °C with continuous agitation at 250 rpm for 2 d, then stored at − 80 °C until use as oligomeric α-synuclein. Monomeric or oligomeric α-synuclein was mixed with 20 μM thioflavin T in 5x assay buffer (250 mM glycine, pH 8.5) in a final volume of 200 μl, and fluorescence was measured at 482 nm with excitation at 446 nm using a PerkinElmer Vitor3 multiplate reader (PerkinElmer, Waltham, MA, USA). Also, a 20 μl aliquot of α-synuclein was adsorbed onto carbon-coated copper grid and air-dried for 10 min. After negative staining with 2% uranyl acetate for another 10 min, α-synuclein was observed with an electron microscope (SIGMA500, Zeiss, Germany).
Plasmid constructs
Plasmid DNA for C-terminal 3xMyc-tagged wild-type LRRK2, LRRK2-G2019S and LRRK2-D1994A, and FLAG-tagged LRRK2 fragments (L1–L4) were kind gifts from Professor Eun-hye Joe (Ajou University, Korea). Myc-tagged mouse SERCA was purchased from Origene (Rockville, MD, USA). pCMV-G-CEPIA1er was a gift from Masamitsu Iino (Addgene plasmid #58215); CMV-R-GECO1mito was a gift from Robert Campbell (Addgene plasmid #46021); pEF-myc-ER-E2-Crimson was a gift from Benjamin Glick (Addgene plasmid #38770); and mito-PAGFP was a gift from Richard Youle (Addgene plasmid #23348).
Subcellular fractionation
HEK293T cells were co-transfected with siRNA targeting the 3′-UTR region of LRRK2 and 3xMyc-tagged wild-type LRRK2 or G2019S-mutated LRRK2. After 24 h, cells were treated with α-synuclein for 24 h. ER, mitochondria, and MAM were isolated from HEK293T cells following the previously described protocols [
64] with minor modifications. Briefly, HEK293T cells at ~ 90–100% confluence were harvested from 25 dishes (10 cm) and homogenized in isolation buffer (225 mM mannitol, 75 mM sucrose, 0.1 mM EGTA, 30 mM Tris-HCl pH 7.4) using a Dounce tissue grinder (Wheaton, Millville, NJ, USA). Nuclei and debris were removed by centrifuging the homogenate twice at 600×g for 10 min, after which the collected supernatant was centrifuged for 15 min at 8000×g. The resulting pellet was collected as the crude mitochondrial fraction. The supernatant was centrifuged at 20,000 g for 1 h, then again at 100,000 g for 1 h, after which the resulting pellet was resuspended as the ER fraction. The supernatant was kept as the cytosolic fraction. For pure mitochondria and MAM fractions, the crude mitochondrial pellet was resuspended in 2 ml mitochondrial resuspension buffer (MRB; 250 mM mannitol, 5 mM HEPES pH 7.4, 0.5 mM EGTA), layered over 30% Percoll medium in a centrifuge tube, and centrifuged at 95,000×g for 30 min. The lower layer (pure mitochondria) and intermediate layer (MAM) between the light membrane and pure mitochondria fractions were then collected. The pure mitochondria fraction was diluted in MRB buffer and further centrifuged at 10,000 g, after which the pellet was resuspended in 300 μl of MRB buffer. The MAM fraction was diluted 10x in MRB buffer and centrifuged at 100,000×g for 1 h, after which the pellet was resuspended in a small volume of MRB buffer.
Time-lapse Ca2+ imaging
Astrocytes isolated from non-Tg or LRRK2-GS mice were plated on glass-bottom dishes (Nest Scientific, China) and then treated with α-synuclein or tunicamycin (0.1 μg/ml) for 24 h. ER Ca2+ release into the cytosol was measured in cells loaded with 5 μM Fluo4-AM (Invitrogen, Carlsbad, CA, USA) in HBSS containing 0.05% Pluronic F-127 (Thermo Fisher Scientific, Waltham, MA, USA) at 37 °C. After 20 min, excess Fluo4-AM was removed by washing cells twice, and then the wash buffer was replaced with HBSS without Ca2+ and Mg2+. Cells were stimulated with 500 μM histamine (Sigma) and imaged under a confocal microscope (TCS, DMi8; Leica, Germany) equipped with a 40x objective (NA 1.10) in a chamber with temperature and CO2 control and a heated stage (LCI, Korea). Images were captured at a rate of one frame per 1 or 2 s. For evaluate Ca2+ dynamics in ER and mitochondria, transfection was performed by incubating astrocytes with a mixture of 1 μg of G-CEPIA1er and R-GECO1mito, Lipofectamine 2000 (Life Technologies) for 5 h. After incubation, the medium was changed and the cells were incubated at 37 °C in 5% CO2. After 24 h, the medium was replaced with HBSS without Ca2+ and Mg2+, and then cells were stimulated with histamine and imaged with a confocal microscope equipped with a 40x objective (NA 1.10). Images were captured at a rate of one frame per 1 or 2 s with the following excitation/emission spectra: G-CEPIA1er, 488 nm/500–550 nm; R-GECO1mito, 552 nm/560–800 nm. After all Ca2+ intensities were corrected for background subtraction, ΔF values were calculated from (F–F0). F0 values, used for normalization, were defined by averaging 10 frames before stimulation.
Immunoprecipitation and nanoLC-MS/MS
HEK293T cells were co-transfected with siRNA targeting the 3′-UTR region of LRRK2 and 3xMyc-tagged wild-type LRRK2 or G2019S-mutated LRRK2. After 24 h, cells were treated with α-synuclein for 24 h and then lysates were prepared using a modified RIPA buffer. Lysates (1 mg) were immunoprecipitated with anti-Myc antibody-conjugated magnetic beads (Invitrogen). Immunoprecipitated proteins were digested with trypsin and quantified on a nanoLC-MS/MS platform. The raw MS files were analyzed and searched against a human protein database based on the species of the sample using Proteome Discoverer 2.0 (Thermo Fisher Scientific). The parameters were set as follows: protein modifications, carbamidomethylation (C) (fixed) and oxidation (M) (variable); enzyme specificity, trypsin; maximum missed cleavages, 2; precursor ion mass tolerance, 10 ppm; and MS/MS tolerance, 0.6 Da. To ensure high confidence identifications, peptide-to-spectrum match (PSM), peptides, and proteins were filtered at a less than 1% false discovery rate (FDR). Three independent replicates for each experimental condition were carried out to control for intrasample variation. In total, 2831 proteins were identified for this project, and were divided into two categories of relative quantitation. Those with a quantitative ratio greater than 1.5 were considered up-regulated, whereas those with a quantitative ratio less than 0.67 were considered down-regulated (Creative Proteomics, Shirley, NY, USA). Protein interactions identified in this study have been submitted to the IMEx (http://www.imexconsortium.org) consortium through IntAct [
34] and assigned the identifier IM-26684.
Cytoscape analysis of LRRK2-interacting proteins
LRRK2-interacting proteins identified by nanoLC-MS/MS were analyzed using the ClueGo Cytoscape plugin with Cellular Component and Biological Process function Gene Ontology annotations. The groups consist of nodes (cellular component) connected to reflect functional relationships between nodes. Each functional group is annotated in colored type.
Analysis of mitochondrial morphology
For visualization of mitochondria, cells were stained with 250 nM MitoTracker Red CMXRos (Invitrogen) for 20 min at 37 °C. Mitochondrial morphology subtypes were quantified using an automated classification system, according to Peng et al. [
39]. After semi-automated segmentation of cell micrographs, mitochondria were classified into six distinct subtypes—small globe, swollen globe, straight tubule, twisting tubule, branch tubule and loop—using automated classification software. The proportion of tubular mitochondria was calculated by totaling straight tubule, twisting tubule, and branch tubule mitochondrial populations. The proportion of fragmented tubular mitochondria was calculated based on small globe mitochondrial populations. High-powered (400 x) fields of micrographs from three independent areas per group were analyzed, and about 200–300 mitochondria from 5 cells were used in calculations.
Tetramethylrhodamine methyl ester imaging
For tetramethylrhodamine methyl ester (TMRM) imaging, astrocytes isolated from non-Tg or LRRK2-GS mice were plated on glass-bottom dishes (Nest Scientific), and then treated with α-synuclein. After 24 h, the medium was replaced with HBSS supplemented with 10 nM TMRM (Invitrogen), and cells were incubated for 30 min at 37 °C. Cells were imaged with a confocal microscope (TCS, DMi8, Leica) equipped with a 40x objective (1.10 NA) in a chamber with temperature and CO2 control and a heated stage. Mitochondrial membrane potential was dissipated by treating cells with 10 μM carbonyl cyanide m-chlorophenyl hydrazone (CCCP; Sigma).
Flow cytometry
Astrocytes isolated from non-Tg and LRRK2-GS mice were treated with α-synuclein for 24–48 h and then stained with the superoxide indicator, MitoSOX Red (5 μM; Invitrogen), for 30 min at 37 °C. Cells were treated with trypsin and analyzed by flow cytometry using a FACS Canto II flow cytometer (BD Biosciences, San Jose, CA, USA).
Immunofluorescence
Astrocytes isolated from non-Tg and LRRK2-GS mice were treated with α-synuclein for 24 h and then fixed with 4% paraformaldehyde and then permeabilized with 0.25% triton for 10 min. Cells were washed three times in PBS and blocked with 1% BSA for 1 h. Cells were incubated with primary antibodies overnight at 4 °C. Washed cells were incubated with fluorescent secondary antibody (Invitrogen) for 1 h at room temperature. Cover slips were mounted with Vectashield containing DAPI (Vector Laboratories) and viewed on confocal microscope confocal microscope (TCS, DMi8; Leica).
TUNEL assay
TUNEL assays were carried out using a commercial kit according to the manufacturer’s instructions (Invitrogen). Briefly, cells and proteinase K-treated organotypic slices were rinsed with phosphate-buffered saline (PBS), and then incubated in 1x equilibration buffer for 10 min. Thereafter, samples were incubated with terminal deoxynucleotidyl transferase (TdT) for 1 h at 37 °C, blocked with stop/wash buffer, and incubated with peroxidase antibody for 30 min at room temperature. The images were taken from layer 1 of the cortex in brain slices (n = 5 slices from 3 mouse) and the percentage of TUNEL-positive cells was determined in at least 10 optical fields.
Proximity ligation assay (PLA)
Fixed cells were stained with the indicated rabbit and mouse antibodies. Duolink-PLA (Olink Bioscience, Sweden) procedures were performed according to the manufacturer’s instructions. Each discrete red spot represents a protein-protein complex (radius < 40 nm).
Proteinase protection assay
ER-enriched fractions from non-Tg and LRRK2-GS astrocytes were prepared using the method described above (refer to Subcellular fractionation). For proteinase protection assays, ER-enriched fractions were diluted in homogenization buffer at a protein concentration of 0.5 μg/μl and incubated in the absence or presence of 150 μg/ml Proteinase K (Invitrogen) for 30 min. The reaction was terminated by adding phenylmethylsulfonyl fluoride (PMSF) to a final concentration of 5 mM and incubating on ice for 5 min. Subsequently, an equal volume of 2x SDS sample buffer was added and protein in the samples was denatured by incubation at 100 °C for 5 min. For Western blot analysis, samples were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS/PAGE). Digestion was monitored using endogenous calnexin and BiP as controls.
Synthesis and transfection of siRNA
siRNA duplex oligonucleotides were chemically synthesized by Bioneer (Daejeon, Korea) and Santa Cruz (Santa Cruz, CA). Details of the siRNA sequences are described in Additional file
2: Table S2. Confluent astrocytes were transfected with siRNA oligonucleotides (50 μM) using Lipofectamine RNAiMax reagent (Invitrogen) according to the manufacturer’s instructions. All assays were performed at least 48 h after RNAi transfection.
Western blotting
Cells were lysed with RIPA buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 2 mM EDTA pH 8.0) supplemented with a protease inhibitor cocktail (GenDEPOT, Barker, TX, USA) at 4 °C for 30 min. Samples were separated by SDS-PAGE and transferred to nitrocellulose membranes. Membranes were incubated with primary antibodies (Additional file
2: Table S4) and horseradish peroxidase (HRP)-conjugated secondary antibodies, and immunoreactive proteins were visualized using an enhanced chemiluminescence system (Ab Frontier, Seoul, Korea).
Real time quantitative reverse transcription-polymerase chain reaction (RT-qPCR) analysis
Total RNA was isolated using RNAiso Plus (TaKaRa, Japan), and cDNA was synthesized using avian myeloblastosis virus reverse transcriptase (New England Biolabs, Ipswich, MA, USA) and oligo (dT) primers (Promega, Madison, WI, USA), according to the manufacturers’ instructions. For qPCR, amplification reactions were performed using a Thermal Cycler Dice Real-Time System (TaKaRa) with SYBR Premix Ex Taq master mix (TaKaRa) according to the manufacturer’s instructions. The primers used for qPCR (Bioneer) are described in Additional file
2: Table S3.
Statistical analysis
The significance of differences between groups was determined using Student’s t-test. A p-value of 0.05 was considered significant. Values are presented as means ± standard deviation (SD).
Discussion
In the current study, we demonstrated that LRRK2-GS acts through SERCA inactivation to trigger ER stress in α-synuclein treated-brain astrocytes, leading to cell death. SERCA is inactivated by its direct interaction with LRRK2-GS, which is translocated to the ER after its release from 14 to 3-3 s. ER-translocated LRRK2-GS persistently maintains the SERCA–PLN complex, thereby disrupting SERCA function and causing ER Ca2+ depletion. Furthermore, LRRK2-GS induced the formation of MAMs and caused Ca2+ overload in mitochondria, leading to mitochondrial dysfunction. Intriguingly, this phenomenon was observed in skeletal muscle cells, another cell type in which SERCA function is prominent.
In most cell types, the ER is the largest intracellular organelle and extends throughout the cytosol. In addition to its role in storing, modifying and transporting newly synthesized proteins, the ER is a high-capacity reservoir for intracellular Ca
2+. SERCAs are responsible for maintenance of the micro- to millimolar Ca
2+ ion concentrations within the ER of eukaryotic cells. Because ER-resident chaperones like CRT, BiP and GRP94 require high Ca
2+ concentrations for their protein-folding activity, the failure of ER Ca
2+ homeostasis resulting from SERCA malfunction brings about an imbalance between the capacity of the protein processing machinery and the amount of unfolded proteins accumulated in the ER, leading to ER-stress–mediated apoptosis [
24]. Moreover, inappropriate Ca
2+ flux from the ER to mitochondria through MAMs impairs several mitochondrial processes. Basal Ca
2+ oscillations drive the mitochondrial metabolism necessary for the production of ATP and mitochondrial substrates used in anabolic processes. In contrast, mitochondrial Ca
2+ overload can cause cell death [
19]. Specifically, this latter study showed that inhibition of SERCA with thapsigargin efficiently killed BAX/BAK
−/− mouse embryo fibroblasts (MEFs) by inducing mitochondrial Ca
2+ overload and opening of the mPTP [
15]. Recent studies have reported that dysregulation of Ca
2+ flux through inactivation of SERCA is involved in several neurological diseases, such as Sandhoff disease and Niemann-Pick A disease [
38]. Using a neuropathic pain model, Gemes and colleagues found that spinal nerve ligation causes a loss of SERCA that results in depletion of ER Ca
2+, and suggested that this may trigger a UPR [
11]. These observations highlight the importance of seeking approaches for restoring appropriate regulation of SERCA function in the treatment of various diseases. In the present study, we showed that LRRK2-GS directly binds and regulates SERCA. This LRRK2-GS–SERCA interaction maintains SERCA–PLN interactions, leading to malfunction of SERCA in the ER (Figs.
3 and
4). PLN is profusely expressed in muscle cells and its role in these cells is well known, but the roles and expression patterns of PLN in the brain are largely unknown. One study reported that PLN is not expressed in neurons [
40], whereas another suggested that Mn
2+ induced the expression of PLN transcripts in astrocytes [
54]. In the current study, we detected PLN expression in both non-Tg and LRRK2-GS astrocytes, but not in neurons (Additional file
2: Figure S5), suggesting that the effect of LRRK2-GS on SERCA may originate from differential expression of PLN. Results obtained in C2 myoblasts further support the idea that this LRRK2-GS mechanism of action is applicable to other cells that express PLN (Fig.
7). Further studies are needed to elucidate whether this is a global or context-specific phenomenon.
The mechanisms that lead to ER dysfunction in PD remain incompletely understood. Moreover, the link between ER dysfunction and PD-related genes, specifically LRRK2, is just starting to be uncovered. Mutated LRRK2 is indirectly involved in ER dysfunction in various ways. First, inhibition of autophagy by mutant LRRK2 might trigger ER dysfunction, given that impairment of autophagy, particularly chaperone-mediated autophagy (CMA), is a reported feature of mutant LRRK2 animal models [
35,
61]. Accordingly, impaired autophagy could be a contributing factor to increased α-synuclein aggregation, which initiates aggregated α-synuclein-dependent ER dysfunction, thereby exacerbating the increase in autophagic cargo and further inhibiting CMA. Consequently, LRRK2 mutations further aggravate ER dysfunction. Secondly, perturbation of endosomal trafficking by mutant LRRK2 imposes a burden on ER. LRRK2 has been shown to interact with several Rab family proteins, which are important regulators of intracellular vesicular trafficking. Mutant LRRK2 phosphorylates Rab, probably leading to altered interactions with downstream effectors of Rab, as well as perturbations in endosome-to-lysosome trafficking [
26,
59]. That means α-synuclein failed to be degraded by lysosomes in mutant LRRK2, adding further pressure on the ER that ultimately leads to ER dysfunction. Lastly, mitochondrial dysfunction caused by mutant LRRK2 impairs physiological functions of the ER. Mutant LRRK2 induces the expression of Bim, a pro-apoptotic BH3-only family member, by phosphorylating the transcription factor FoxO [
16]. Many of these BH3-only proteins affect ER Ca
2+ homeostasis by binding to the IP
3R and/or changing its phosphorylation status, thereby altering the Ca
2+-flux properties of the channel [
47]. Moreover, PD-linked LRRK2 mutations increase mitochondrial Ca
2+ uptake in cortical neurons in association with increased expression of the mitochondrial Ca
2+ uniporter (MCU) [
63]. As shown by our MS/MS analysis in the current study, LRRK2-GS also interacts with proteins involved in mitochondria membrane organization and mitochondrial transport, suggesting that LRRK2 may contribute to mitochondria function in astrocytes (Additional file
1: Table S1 Additional file
2: Figure S3). Accordingly, the effect of mutant LRRK2 on mitochondria may induce de novo ER stress or aggravate existing ER dysfunction.
In this study, we demonstrated the molecular mechanism by which G2019S-mutated LRRK2 induces ER stress and subsequent cell death. By incorporating previously unidentified LRRK2-interacting components, this study fills an important gap in our understanding of the relationship between LRRK2-GS and PD pathogenesis. Specifically, our study suggests a plausible model that links LRRK2-GS to pathophysiological ER stress in PD. In addition, novel aspects of this work related to mutated LRRK2 in muscle cells will help shed light on how mutated LRRK2 leads to abnormal phenotypes in peripheral tissues. A better understanding of LRRK2 biology in the context of ER stress could impact ongoing efforts to establish the action mechanism of LRRK2 in both PD and peripheral organs.