Identification of ciliary neurotrophic factor receptor alpha as a mediator of neurotoxicity induced by alpha-synuclein

Abstract

Accumulating evidence suggests that extracellular alpha-synuclein (eSNCA) plays an important role in the pathogenesis of Parkinson's disease or related synucleinopathies by inducing neurotoxicity directly or indirectly via microglial or astroglial activation. However, the mechanisms by which this occurs remain to be characterized. To explore these mechanisms, we combined three biochemical techniques - stable isotope labeling of amino acid in cell cultures (SILAC), biotin labeling of plasma membrane proteins followed by affinity purification, and analysis of unique proteins binding to SNCA peptides on membrane arrays. The SILAC proteomic analysis identified 457 proteins, of which, 245 or 172 proteins belonged to membrane or membrane associated proteins, depending on the various bioinformatics tools used for interpretation. In dopamine neuronal cells treated with eSNCA, the levels of 86 membrane proteins were increased and 35 were decreased compared with untreated cells. In peptide array analysis, 127 proteins were identified as possibly interacting with eSNCA. Of those, seven proteins were overlapped with the membrane proteins that displayed alterations in relative abundance after eSNCA treatment. One was ciliary neurotrophic factor receptor, which appeared to modulate eSNCA-mediated neurotoxicity via mechanisms related to JAK1/STAT3 signaling but independent of eSNCA endocytosis.

Fig. 1. Affinity enrichment of plasma membrane proteins

A) MES cells, grown in media at 37 °C, were washed with ice-cold PBS (pH 8.0) three times before incubating with sulfo-NHS-LC-biotin (0.5mg/ml in pH 8.0 PBS) for 20 min at 4°C. The unreacted sulfo-NHS-LC-biotin was quenched (at 37 °C for 10 min) by 100 mM glycine solution, followed by washing cells three times with ice-cold PBS. Biotinylated MES cells were visualized using avidin-linked fluorescent that highlights the plasma membrane (a magnified view shown as an insert at the right upper corner). Background staining could be due to leaking of biotin into the cells. Scale bar 27.5 μm. B) Biotinylated MES cells were fractionated with sucrose gradient ultracentrifugation before obtaining affinity-enriched membrane fraction (Dynabeads streptavidin at 4 °C for 1 hr). A total of 6 μg of whole–cell lysate (WC), crude membrane fraction (CM), or affinity-purified plasma membrane (PM) proteins were separated by 8–16% SDS-PAGE, transferred to a PVDF membrane, and blotted with an antibody against organelle-specific proteins: anti-annexin II for plasma membrane; anti-cytochrome C for mitochondrion; anti-β-actin for cytosol. C) Total amount of protein loaded for each sample was similar as revealed by silver staining.

Fig. 2. Identification of plasma membrane proteins of MES cells

A) MES Cells were treated with human pre-aged eSNCA for 1, 3, or 6 hrs, washed extensively with PBS. The extent of internalization of human eSNCA was measured using Western blotting with β-actin as an internal control. Internalized eSNCA appeared as early as 1 hr after incubation. Notably, endogenous SNCA cannot be recognized with this antibody, i.e., rat-mouse SNCA level does not influence the level of internalized human eSNCA. B) GO analysis and annotation of total proteins identified by 2 or more peptides (proteins identified with high confidence). Proteins are classified according to their participation in specific biological functions (B1) and their cellular components (B2), respectively.

Fig. 3. Detection of biotinylated membrane proteins bound to SNCA peptide arrays

Biotinylated membrane proteins were enriched and the lysate was incubated with an SNCA peptide array membrane. After extensive washing to remove non-specific binding, the positive spots were detected using avidin-conjugated HRP, followed by ECL reagent development. Positive spots are marked with arrows.

Fig. 4. Interaction between SNCA and CNTFR-α by immunostaining and immunoprecipitation

A) MES cells, seeded in a 4-well chamber at 0.05×10⁶/well, were treated with eSNCA at 250 nM for 1 hr. Cells were stained with anti-CNTFR-α (red-A1) and anti-SNCA (green-A2) antibody, followed by examination with confocal microscopy. The nuclei are highlighted by DAPI (blue-A3). Note: SNCA was co-localized with CNTFR-α after 1 hr treatment (yellow color of the merged image-A4). Scale bar 5.4 μm. B) The protein complex of interest was isolated from MES cells homogenate after eSNCA treatment. Co-immunoprecipitation analysis using magnetic beads conjugated with either SNCA (top panel) or CNTFR-α (bottom panel) with subsequent pull-down revealed a noticeable protein-protein association between SNCA and CNTFR-α. Input represents original materials. CTL represents normal mouse or goat IgG control.

Fig. 5. Effect of knocking down CNTFR-α in MES cells on the loss of neurites

A) MES cells were transfected with CNTFR-α siRNA and expression levels of CNTFR-α were analyzed by Western blotting with an CNTFR-α antibody 72 hrs after gene manipulation, demonstrating that CNTFR-α siRNA effectively inhibited CNTFR-α expression (*p<0.05, compared to non-transfected control [CTL]/non-sense [NS] siRNA-transfected groups with β-actin as a loading control). B) Non-transfected MES cells or cells transfected with CNTFR-α siRNA or NS siRNA were seeded in 4-well chambers and incubated for 72 hrs. Cells were then treated with rotenone at 5 nM for 6 hrs, aggregated eSNCA for 1 hr, or a combination of rotenone and SNCA. After fixing and staining with an anti-MAP2 antibody, images of five randomly selected fields were then captured using a laser scanning confocal microscope and three wells were counted in each experimental condition. The length of neurites on MAP2+ cells and number of branching points of six cells in each field were quantified using Neurolucida software (version 8.0, MicroBrightField, Williston, VT, USA) by an observer blind to the experimental settings. Data are means ± SEM of at least three independent determinations (* p<0.05). Notably, cells transfected with NS siRNA or CNTFR siRNA and treated with rotenone or SNCA alone were also assessed, demonstrating no significant differences in terms of neurite out-growth when compared to non-transfected treated (Rot or SNCA) controls, or untreated (CNTFR siRNA-transfected or not) controls (data not shown). CTL represents non-transfected un-treated controls.

Fig. 6. JAK1/STAT3 signaling pathway and eSNCA-induced neurotoxicity via CNTFR-α

A) MES cells were transfected with CNTFR-α siRNA or non-sense [NS] siRNA before treatment with eSNCA. Untreated cells (CTL) and non-transfected treated cells were used as controls. The extent of internalization of human eSNCA in cells post-treatment 1 hr was measured by Western blotting analysis using an antibody specific to human α-synuclein. The internalization of human eSNCA was not blocked by knocking down CNTFR-α. B) MES cells were transfected with CNTFR-α siRNA or NS siRNA before treatment with eSNCA. The amount of JAK1 in cells was measured by Western blotting analysis. A significantly decreased JAK1 expression in cells was found after CNTFR-α knock-down (*p<0.05, densitometry values are represented as JAK1/β-actin ratio and means ± SD from three independent experiments, compared to treated non-transfected control [Treated] and non-sense [NS] siRNA-transfected groups; untreated NS or CNTFR siRNA-transfected groups were similar to the untreated control [CTL] group [data not shown]; β-actin was used as a loading control). C) The nuclear translocation of STAT3 in MES cells was visualized after eSNCA treatment with or without CNTFR-α siRNA transfection using immunostaining. While STAT3 was translocated to the nuclear compartment after eSNCA treatment, the knock-down of CNTFR-α appeared to inhibit the translocation. Scale bar, 9 μm.