α-Synuclein amyloid fibril directly binds to LC3B and suppresses SQSTM1/p62-mediated selective autophagy

Dear Editor,

Parkinson’s disease (PD) is a progressive neurodegenerative disorder, which manifests through the abnormal accumulation of pathological amyloid fibrils composed of α-synuclein (α-syn) into Lewy bodies and the deterioration of dopaminergic neurons in the substantia nigra.^1–4 In addition to α-syn fibrillar aggregation, the disruption of selective autophagy is also tightly linked to the pathogenesis of PD.^5,6 The co-localization of LC3B, the key autophagosome protein in selective autophagy,⁷ and α-syn in the Lewy bodies of PD patients’ brains points towards a potential role of α-syn in modulating selective autophagy.⁸ Yet, the interplay between them has not been mechanistically elucidated.

First, we employed biolayer interferometry (BLI) to determine the binding affinity between LC3B and α-syn, in its either monomeric or fibrillar form. The BLI data revealed a significant difference in binding affinities, with the equilibrium dissociation constant (K_D) for the interaction between LC3B and α-syn preformed fibrils (α-syn PFFs) being three orders of magnitude stronger than that for LC3B and α-syn monomer (Fig. 1a). Further validation through enzyme-linked immunosorbent assay (ELISA) confirmed LC3B’s preferential affinity for α-syn fibrils over the monomers (Supplementary information, Fig. S1a). Consistent with the biophysical measurements, we visualized direct attachment of the His-tagged LC3B to the surface of α-syn fibrils through immunoelectron microscopy (Fig. 1b), whereas control experiments without LC3B showed minimal nanogold particle adherence to α-syn fibrils (Supplementary information, Fig. S1b). These findings underscore the direct and selective interaction between LC3B and α-syn fibrils, highlighting a preferential LC3B binding to the fibrillar over the monomeric form.

Fig. 1. α-Syn amyloid fibril directly binds LC3B and suppresses SQSTM1/p62-mediated selective autophagy.

a Kinetic binding curves of LC3B with α-syn PFFs (left) and α-syn monomer (right) measured by BLI assay. b Transmission electron microscope images of α-syn fibrils incubated with His-tagged LC3B and probed by anti-His nanogolds. The fibrils are zoomed in on the right. Red arrows indicate the attachment of nanogolds on α-syn fibrils. Scale bars, 50 nm. c, d 2D ¹H-¹⁵N HSQC spectra of ¹⁵N-α-syn monomer titrated by series concentrations of LC3B. The crosspeaks of the residues at the C-terminal region of α-syn with significant CSDs are highlighted with black boxes (c) and enlarged in d. e Residue-specific CSDs of ¹⁵N-α-syn monomer titrated by LC3B. The primary sequence of α-syn C-terminal region (residues 110–140) is shown on the top, and the residues with CSDs > 0.03 ppm (molar ratio of 1:5) are highlighted in red. f Residue-specific CSDs of ¹⁵N-LC3B titrated by α-syn. g Electrostatic surface of the representative complex structure of LC3B–α-syn_101–140 predicted by AlphaFold-Multimer. The complex interface is zoomed in on the right with LC3B residues labeled in blue and α-syn residues labeled in red. h The comparison of the binding sites of α-syn and p62 LIR on LC3B. i Schematic diagram of the NMR competition experiment. The chemical shift of ¹⁵N-α-syn caused by LC3B decreases with the addition of p62 LIR due to the competitive binding. j Residue-specific CSDs of ¹⁵N-α-syn titrated by LC3B and p62 LIR peptide. The colored arrows show the changes of chemical shifts of α-syn caused by LC3B and p62 LIR peptide. k Zoomed-in images of crosspeaks of representative residues in NMR competition experiment with colored arrows showing the moving directions of ¹⁵N-α-syn chemical shifts caused by LC3B and p62 LIR peptide. l Representative western blot images and quantifications of the autophagy flux of p62 in HeLa cells after treatment of the indicated concentrations of α-syn PFFs vs the vehicle control (0 nM). m Representative fluorescence microscopic images showing the ratio of p62 inside the lysosomes vs p62 outside the lysosomes detected by EGFP-mCherry-p62 dual fluorescence reporting system in HEK293T-WT-α-syn cells. Scale bar, 5 μm. The quantification of ratios is shown on the right. n Similar to l, but in SHSY-5Y-WT-α-syn cells transfected with α-syn PFFs (800 nM). The level of K63-ubiquitylated proteins was tested instead of p62 as an additional indicator of selective autophagy. Data in l, m and n are means ± SEM. n indicates biological replicates. Unpaired t-tests. *P < 0.05, **P < 0.01, and ***P < 0.001.

Next, we utilized solution nuclear magnetic resonance (NMR) spectroscopy to identify the key regions within the α-syn responsible for LC3B binding. Analysis through two-dimensional (2D) ¹H-¹⁵N heteronuclear single-quantum coherence (HSQC) spectra revealed that titrating LC3B to ¹⁵N-labeled α-syn monomer led to significant chemical shift deviations (CSDs) and a reduction in signal intensity for residues primarily localized in the C-terminal region of α-syn, particularly residues 110–140, in a LC3B concentration-dependent manner (Fig. 1c–e; Supplementary information, Fig. S2a). The observation of significant CSDs in negatively charged acidic residues within the C-terminal region of α-syn suggests a pivotal role for electrostatic interactions in the binding process.

Previous studies have shown that the C-terminal region of α-syn maintains a similar conformational ensemble structure in both its monomeric and fibrillar forms,^9,10 suggesting that the interaction between LC3B and α-syn fibril is likely mediated by the C-terminal region. To test this, we generated a C-terminal truncated variant of α-syn (referred to as α-syn₁₀₀) and produced fibrils from α-syn₁₀₀ that shared morphological characteristics with wild-type (WT) α-syn fibrils (Supplementary information, Fig. S2b). BLI assays demonstrated that LC3B binds to neither the monomeric nor the fibrillar form of α-syn₁₀₀ (Supplementary information, Fig. S2c), confirming the indispensable role of the C-terminal region in mediating LC3B’s binding to α-syn monomers and fibrils. Previously, we had demonstrated that the C-terminal region is intermittently concealed in the monomeric form but becomes fully exposed and well aligned along the surface of the fibril core at high densities (with hundreds of C-terminal segments per fibril).⁹ This might well explain the preference of LC3B for binding the fibrillar rather than the monomeric form of α-syn.

We next sought to pinpoint the specific interface on LC3B that interacts with α-syn. By employing 2D ¹H-¹⁵N HSQC spectroscopy on ¹⁵N-labeled LC3B, we observed significant CSDs and signal intensity drop upon titration of α-syn monomer to ¹⁵N-labeled LC3B (Fig. 1f; Supplementary information, Fig. S3a, b). By mapping residues that exhibited CSDs greater than 0.1 ppm (at a ratio of LC3B:α-syn = 1:2) onto the 3D structural model of LC3B (PDB: 3VTU), we found that residues displaying significant CSDs gather around four positively charged residues (H27/K49/H57/R69), which collectively form a positively charged pocket on the surface of LC3B (Supplementary information, Fig. S3c).

To clarify the necessity of this positive pocket in the interaction, we mutated these four basic residues to acidic ones (H27D/K49E/H57D/R69E), creating a variant referred to as LC3B_mut, which negates LC3B’s positively charged surface. Circular dichroism (CD) analysis verified that these mutations did not alter LC3B_mut’s secondary structure compared to native LC3B (Supplementary information, Fig. S4c). Subsequent titration of ¹⁵N-α-syn with LC3B_mut resulted in negligible CSDs and no notable decrease in signal intensity (Supplementary information, Fig. S4a, b). As predicted, BLI assay demonstrated that these mutations effectively eliminate LC3B’s ability to bind to both α-syn fibrils and monomers (Supplementary information, Fig. S4d).

Additionally, we utilized AlphaFold-Multimer¹¹ to predict LC3B-α-syn complex structure and the result showed that α-syn’s C-terminal region binds to the positively charged surface on LC3B, aligning closely with our in vitro findings from NMR and BLI experiments (Fig. 1g). Overall, our findings elucidate that the interaction between LC3B and α-syn is governed by electrostatic interactions, through the engagement of the negatively charged C-terminal region of α-syn with the positively charged surface on LC3B.

For the other two isoforms of LC3 subfamily, LC3A and LC3C, which share high sequence and structural similarity with LC3B (Supplementary information, Fig. S5a–f), their binding affinities toward α-syn were also determined using BLI. Both LC3A and LC3C could interact with α-syn, and the binding affinities are similar to that of LC3B for α-syn (Supplementary information, Fig. S5c, e). Moreover, LC3A and LC3C show higher affinities to α-syn PFFs than to α-syn monomer, demonstrating that these three LC3 isoforms exhibit similar binding patterns with α-syn. As LC3B is the most well-studied marker protein in autophagy and is expressed in almost all tissues in mammals,¹² we chose LC3B for further study.

SQSTM1/p62 (referred to as p62 hereafter), a key receptor of selective autophagy, serves to identify ubiquitinated substrates and deliver them to LC3B using its LC3-interacting region (LIR).¹³ Notably, structural analysis revealed that the α-syn-binding site on LC3B significantly overlaps with the binding site for the p62 LIR motif (Fig. 1h; Supplementary information, Fig. S6a), suggesting a potential competitive binding scenario between α-syn fibrils and p62. This competition could potentially disrupt the p62–LC3B interaction, impairing selective autophagy and protein homeostasis. To investigate this competition at the residue level, we conducted an NMR titration experiment. Introducing LC3B to ¹⁵N-α-syn monomer followed by the incremental addition of p62 LIR peptide led to a decrease in CSDs induced by LC3B, confirming that p62 LIR peptide’s competitive binding lessens the interaction between LC3B and α-syn (Fig. 1i–k; Supplementary information, Fig. S6b). Subsequently, we performed pull-down assays to verify the impact of α-syn PFFs on the p62–LC3B interaction. Treatment with varying concentrations of α-syn PFFs led to a reduction in p62’s binding to LC3B in a dose-dependent manner (Supplementary information, Fig. S6c). These findings confirmed that α-syn PFFs can interfere with the p62–LC3B binding in vitro.

To assess the potential influence of α-syn PFFs in p62-mediated selective autophagy in cells, we introduced α-syn PFFs into HeLa cells through liposome-mediated transfection over a 48 h period to induce intracellular α-syn fibrillization. The formation of α-syn aggregates specifically resulting from α-syn PFFs was monitored using a homogeneous time-resolved fluorescence (HTRF) assay (Supplementary information, Fig. S7a–d). These cells were then treated with either the autophagy inhibitor bafilomycin A1 (BafA1) or the DMSO control for another 24 h to evaluate autophagic fluxes by measuring the p62 level.¹⁴ Our western blot analysis indicated a rise in the steady-state level of p62 protein and a corresponding decrease in p62 autophagy flux in HeLa cells (Fig. 1l; Supplementary information, Fig. S7a). This phenomenon was also observed in HEK293T cells and SHSY-5Y cells overexpressing the A53T variant of human α-syn (SHSY-5Y-A53T-α-syn), an α-syn variant associated with familial forms of PD (Supplementary information, Fig. S7b, c, e, f). From these observations, we inferred that under pathological conditions, cellular α-syn fibrils might preferentially bind to LC3B, thereby disrupting the p62–LC3B interaction, leading to elevated p62 protein levels and reduced p62-dependent autophagy flux. By using an EGFP-mCherry-p62 dual fluorescence reporting system, we found that α-syn PFF treatment blocked p62 from entering lysosomes (Fig. 1m). Given that K63-linked polyubiquitinated proteins are typically recognized by p62 and subsequently degraded via the LC3B-p62 selective autophagy pathway,¹⁵ we assessed the overall levels of K63-polyubiquitinated proteins in SHSY-5Y cells overexpressing WT α-syn (SHSY-5Y-WT-α-syn) after α-syn PFF transfection. As expected, an increase in K63-ubiquitylated protein levels and a reduction in their autophagy flux were observed (Fig. 1n; Supplementary information, Fig. S7d). We constructed autophagy activation cell models by starvation or treatment with mTOR inhibitors Torin1 or rapamycin, and then tested the inhibitory effect of α-syn PFFs treatment on p62 degradation. In cells with activated autophagy, our western blot analysis indicated an increased steady-state levels of p62 protein following α-syn PFF treatment (Supplementary information, Fig. S8). BafA1 increased the basal level of p62, but significantly reduced the enhancement of p62 levels caused by α-syn PFF treatment, indicating an impaired autophagy flux. In conclusion, in the case of autophagy activation, the competitive binding of α-syn PFFs and p62 to LC3B can also inhibit the LC3B–p62-mediated selective autophagy, resulting in impaired autophagy flux of p62.

To substantiate that the changes in p62 protein levels and autophagic flux in cellular models were directly attributable to the interaction between LC3B and α-syn PFFs, we tested the effects of α-syn₁₀₀ fibrils by introducing α-syn₁₀₀ PFFs into HEK293T cells stably expressing α-syn₁₀₀. The generated cellular α-syn₁₀₀ fibrils generated at these conditions do not interact with LC3B. Western blot analysis showed that the induced α-syn₁₀₀ fibrils did not affect p62 protein levels as the full-length α-syn fibrils did (Supplementary information, Fig. S9a, b). We further established ATG5- or LC3B-knockout HEK293T cell line stably expressing full-length α-syn (denoted as HEK293T-WT-α-syn-ATG5KO or HEK293T-LC3BKO) to confirm that the observed p62 level changes depend on LC3B and autophagy (Supplementary information, Fig. S9c, d), validating that the increased p62 levels result from reduced selective autophagy flux.

In this study, we show that α-syn fibrils may suppress selective autophagy via their direct binding to LC3B. In vitro experiments demonstrated that α-syn fibrils interact with LC3B with a significantly higher affinity compared to the α-syn monomers. α-Syn fibrils utilize their negatively charged C-terminal region to bind the positively charged pocket of LC3B. Importantly, the binding site of α-syn on LC3B considerably overlaps with that of p62, leading to reduced LC3B–p62 interaction via the competitive binding of α-syn fibrils. Furthermore, α-syn fibril treatment in cells reduced the autophagic flux of p62 and K63-ubiquitinated proteins, which are degraded via selective autophagy in cells. This suggests that under pathological conditions, excessive α-syn fibril binding to LC3B might occupy the interface needed for p62 binding, thereby impairing p62-mediated selective autophagy. This could lead to a breakdown in selective autophagy and ultimately disrupts cellular protein balance. Our cellular experiment results showed a modest but significant suppression of selective autophagy by α-syn PFF treatment for a relatively short time (hours), while in the long-term progression of PD (years), α-syn aggregation may have a chronic and accumulative impact on selective autophagy.

Our results highlight a direct link between α-syn aggregation and selective autophagy malfunctioning, revealing to a converging pathological pathway in PD and offering novel insights into potential new therapeutic strategies for the disease.

Author contributions

C.L., B.L, Q.X. and H.W. designed the project. Q.X. performed in vitro study on molecular mechanism of protein interaction. H.W. performed cellular experiments, verified the role of protein interactions in cell models. R.Y. and Z.W. assisted with cellular experiments. Y.T. and S.Z. assisted with NMR experiments. B.S. and D.L. assisted with data analyses. Q.X. and H.W. prepared figures. Q.X., H.W., B.L. and C.L. wrote the manuscript. All the authors were involved in data analyses and contributed to manuscript discussion and editing. All authors reviewed and approved the final version of the manuscript.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41422-024-01022-2.