α-Synuclein and huntingtin exon 1 amyloid fibrils bind laterally to the cellular membrane

Abstract

Fibrillar aggregates involved in neurodegenerative diseases have the ability to spread from one cell to another in a prion-like manner. The underlying molecular mechanisms, in particular the binding mode of the fibrils to cell membranes, are poorly understood. In this work we decipher the modality by which aggregates bind to the cellular membrane, one of the obligatory steps of the propagation cycle. By characterizing the binding properties of aggregates made of α-synuclein or huntingtin exon 1 protein displaying similar composition and structure but different lengths to mammalian cells we demonstrate that in both cases aggregates bind laterally to the cellular membrane, with aggregates extremities displaying little or no role in membrane binding. Lateral binding to artificial liposomes was also observed by transmission electron microscopy. In addition we show that although α-synuclein and huntingtin exon 1 fibrils bind both laterally to the cellular membrane, their mechanisms of interaction differ. Our findings have important implications for the development of future therapeutic tools that aim to block protein aggregates propagation in the brain.

Figure 1. Schematics for the different scenarios of amyloid fibrils binding to biological membranes.

(a) Binding mainly through fibrils extremities. At identical monomer concentration and under non-saturating conditions, short fibrils will bind the cellular membrane to a higher extent than longer fibrils. (b) Lateral binding that does not depend on the number of interacting sites. Long and short fibrils will bind equally well to the cellular membrane. (c) Lateral binding with a minimal number of interacting sites required. Binding becomes dependent on fibrils length as the longer the fibrils are the higher the probability of interacting with the minimal number of interaction sites is. Fibrils of a length that do not meet with the required number of interaction sites will not bind the cellular membrane at all.

Figure 2. Morphologies and length distributions of long and short αSyn and HTTExon1 fibrils.

(a,c) Representative negatively stained TEM of long (red box) and short (blue box) αSyn (a) or HTTExon1 (c) fibrils, freshly sonicated (box with solid lines) or after being incubated for 1 hour with unilamellar vesicles made from brain lipids in Dulbecco’s modified Eagle’s medium (box with dashed lines). Scale bar, 100 nm. (b,d) Length distributions of long (red) and short (blue) αSyn (b) or HTTExon1 (d) fibrils, freshly sonicated (closed circles, solid lines) or after being incubated for 1 hour with unilamellar vesicles made from brain lipids (open circles, dashed lines). The length distributions were obtained by measuring the length of at least 1000 fibrils from 3 to 10 independent experiments by quantitative negatively stained TEM.

Figure 3. Long and short fibrils have similar structures.

(a) ThT fluorescence spectra following binding to fibrils. (b) Proteinase-K degradation patterns of αSyn long (red) and short (blue) fibrils at identical monomeric protein concentrations. (c) Proteinase-K degradation kinetics of SDS-resistant long (red) and short (blue) HTTExon1 fibrils at identical monomeric protein concentrations. Insets, representative SYPRO-stained SDS-PAGE showing the disappearance of fibrillar HTTExon1 trapped within the gel well as a function of time. (d) Proteinase-K degradation pattern of TFA-HFIP solubilized long (red) and short (blue) HTTExon1 fibrils as a function of time. SDS-PAGE were SYPRO-stained. In panels a and c means and standard errors are calculated from 3 independent experiments.

Figure 4. Binding of long and short αSyn and HTTExon1 fibrils to unilamellar vesicles made from brain lipids.

Binding of 50 μM αSyn long (a) and short (b) fibrils and of 50 μM HTTEx1 long (c) and short (d) fibrils to a 5× mass excess of artificial liposomes made from total brain lipids assessed by TEM. Scale bar, 100 nm.

Figure 5. Binding of long and short αSyn and HTTExon1 fibrils to Neuro 2A cells.

(a) Binding of αSyn-ATTO488 long (red) and short (blue) fibrils at increasing concentrations (0.1–5 μM) to Neuro 2A cells assessed by flow cytometry. (b,c) Representative traces of the binding of 1 μM αSyn-ATTO488 long (b) and short (c) fibrils to Neuro 2A cells. (d) Binding of 1 μM αSyn-ATTO488 long fibrils to Neuro 2A cells in the presence of increasing concentrations (0–100 μM) of unlabelled long αSyn fibrils. (e,f) Representative traces of the binding to Neuro 2A cells of αSyn-ATTO488 long fibrils (1 μM) in the presence of 10 (e) or 100 (f) μM unlabelled long αSyn fibrils. (g) Binding of HTTExon1-ATTO488 long and short fibrils at increasing concentrations (0.5–10 μM) to Neuro 2A cells. (h,i) Representative traces of the binding to Neuro 2A cells of HTTExon1-ATTO488 long fibrils (h, 5 μM), and HTTExon1-ATTO488 short fibrils (i, 5 μM). The means percentage of cells with bound fibrils and the associated standard error values, calculated from 3 to 6 independent experiments, are represented in (a,d,g). The horizontal line represents the limit between non fluorescent and fibrils-bound, fluorescent cells in (b,c,e,f,h,i).

Figure 6. Assessment of Neuro 2A cells permeabilization by long and short fibrils by imaging intracellular free Ca²⁺ alterations.

(a–c) Cells loaded with Fluo4-AM were imaged by epifluorescence microscopy after exposure to PBS or PBS + 5 mM EGTA (a), long and short αSyn fibrils (10 μM; b), and long and short HTTExon1 fibrils (10 μM; c). Scale bars, 20 μm. (d) Quantification of the intracellular free Ca²⁺ increase over time, expressed as a fraction of the maximal Ca²⁺ increase observed after addition of ionomycin (10 μM) to the cells. Data are means and associated standard error values calculated from 6 to 10 independent experiments. For the sake of clarity the standard errors are not represented for αSyn and HTTExon1 long and short fibrils in the presence of 5 mM EGTA. For each experiment, fluorescence increase is a mean calculated over ≈50–100 individual cells. The increase in fluorescence measured after addition of PBS + 5 mM EGTA was subtracted for each trace.