α-synuclein-assisted oligomerization of β-amyloid (1-42)

Abstract

Alzheimer’s disease (AD) and Parkinson’s disease (PD) are the two most common neurodegenerative disorders, characterized by aggregation of amyloid polypeptides, β-amyloid (Aβ) and α-synuclein (αS), respectively. Aβ and αS follow similar aggregation pathways, starting from monomers, to soluble toxic oligomeric assemblies, and to insoluble fibrils. Various studies have suggested overlaps in the pathologies of AD and PD, and have shown Aβ-αS interactions. Unfortunately, whether these protein-protein interactions lead to self- and co-assembly of Aβ and αS into oligomers – a potentially toxic synergistic mechanism – is poorly understood. Among the various Aβ isoforms, interactions of Aβ containing 42 amino acids (Aβ (1-42), referred to as Aβ42) with αS are of most direct relevance due to the high aggregation propensity and the strong toxic effect of this Aβ isoform. In this study, we carefully determined molecular consequences of interactions between Aβ42 and αS in their respective monomeric, oligomeric, and fibrillar forms using a comprehensive set of experimental tools. We show that the three αS conformers, namely, monomers, oligomers and fibrils interfered with fibrillization of Aβ42. Specifically, αS monomers and oligomers promoted oligomerization and stabilization of soluble Aβ42, possibly via direct binding or co-assembly, while αS fibrils hindered soluble Aβ42 species from converting into insoluble aggregates by the formation of large oligomers. We also provide evidence that the interactions with αS were mediated by various parts of Aβ42, depending on Aβ42 and αS conformers. Furthermore, we compared similarities and dissimilarities between Aβ42-αS and Aβ40-αS interactions. Overall, the present study provides a comprehensive depiction of the molecular interplay between Aβ42 and αS, providing insight into its synergistic toxic mechanism.

1. INTRODUCTION

Alzheimer’s disease (AD), the most prevalent neurodegenerative disease, is characterized by the loss of cognitive function and onset of dementia [1-3]. AD is pathologically linked to the amyloidogenic aggregation of β-amyloid (Aβ) [2-4], which is predominantly found as either 40 or 42 residue long peptides, referred to as Aβ40 and Aβ42, respectively [1-3]. Aβ peptides are produced by the abnormal cleavage of the transmembrane protein, amyloid precursor protein (APP) [2, 3, 5]. Proteolysis of APP by β-secretase and γ-secretase results in release and accumulation of Aβ peptides in the extracellular space [2, 3], where the intrinsically disordered Aβ monomers can spontaneously aggregate to form “soluble” oligomeric assemblies [1, 3, 4], which can further aggregate into “insoluble” fibrils [1, 6]; of which the amyloid plaques of AD patients are comprised [7]. The soluble oligomers are widely regarded as the most toxic, instigating neuronal damage and leading to dementia [8-13].

Parkinson’s disease is the second most common neurodegenerative disorder, after AD, and commonly diagnosed by motor function symptoms [14]. This disease is characterized by the loss of dopamine neurons in the substantia nigra and prevalence of intraneuronal Lewy bodies (LB), which are composed of aggregates of the 140-resdiue amyloid protein, α-synuclein (αS) [15-18]. Analogous to Aβ, αS follows an amyloidogenic aggregation pathway that is associated with the pathology of PD [19]. αS oligomers rich in β-sheet structure are the main toxic species responsible for the propagation of PD, for example, by disrupting the integrity and permeability of the neuronal membrane [20-23].

Recently, there have been renewed interest in uncovering the connections between AD and PD. A correlation first arose due to similarities in their pathologies and the discovery of the fragment of αS, known as the non-amyloid-β component (NAC), in the amyloid plaques of AD patients [24]. Additionally, many AD patients have exhibited αS Lewy bodies, a disease referred to as dementia with Lewy bodies (DLB) [25]. Patients diagnosed with PD with dementia (PDD) can also accumulate Aβ plaques in the brain [26-28]. Previous in vivo studies with human brain tissues and transgenic (Tg) mice showed greater neuronal degeneration and more severe deficits in cognition and motor skill when Aβ and αS were both present [29-32], supporting the reality of synergistic interactions between Aβ and αS [33-38]. Moreover, strong evidence in the literature has suggested that the synergistic toxic effects might result from oligomerization driven by Aβ-αS interactions. For example, intracerebral injections of αS-containing brain extracts into AD mice inhibit Aβ deposition into fibrillar plaques while increasing soluble oligomers of Aβ and αS [39]. Moreover, the level of soluble αS oligomers is higher and amyloid plaque load is lower in human brain with AD and PD co-pathologies than in AD alone [40, 41]. Aβ and αS co-expression in double Tg mice also increases αS oligomerization compared to single Tg mice, without increasing Aβ fibrillar plaques [32, 41-43]. Taken altogether, these cases suggest that Aβ and αS can eventually interact in the brain both intracellularly and extracellularly, with findings that intraneuronal αS can be excreted into the extracellular space [44] and extracelluar Aβ can be uptaken within the neuron [45].

Earlier in vitro studies demonstrated protein-protein interactions between Aβ and αS at the molecular level [41, 42, 46-52], showing that Aβ and αS can enhance each other’s aggregation [30, 42, 49, 50, 53-55]. While informative, most of these studies provided limited insight into consequences of Aβ-αS interactions at the molecular level, because aggregation was poorly defined with no distinction between oligomerization and fibrilization. Moreover, amyloid aggregation was also often characterized under denaturing conditions (e.g., with sodium dodecyl sulfate (SDS)), which can introduce undesired artifacts on aggregation states [47, 56]. Recently, in our in vitro study with Aβ40 and αS, we examined protein-protein interactions between Aβ40 and αS under non-denaturing conditions [57]. The study showed that Aβ40 fibrillization was inhibited by αS monomers and oligomers, which however, promoted Aβ40 oligomerization and stabilized preformed Aβ40 oligomers [57]. The study also provided evidence that the Aβ40-αS interactions occurred primarily on the C-terminus of Aβ40. This is noteworthy since Aβ42, the more aggregation-prone and toxic isoform of Aβ, differs from Aβ40 by the two C-terminal amino acids (ILe-Ala). Despite a stronger aggregation propensity and more significant toxic effects of Aβ42, relative to Aβ40 [1-3, 58], molecular consequences of Aβ42-αS interactions have yet to be determined.

In the present study, we examined aggregation-state specific Aβ42-αS interactions and their molecular outcomes in terms of aggregation, particularly oligomerization. For this examination, we prepared samples of Aβ42 and αS in their monomeric, oligomeric and fibrillar forms, which were then subjected to a set of experimental tools to interrogate Aβ42-αS interactions in their amyloid assembly. Our results demonstrate the role of αS in promoting Aβ42 oligomerization possibly via direct binding or co-assembly and inhibiting Aβ42 fibrillization, similar to the previously reported effects of αS monomers and oligomers on Aβ40. Interestingly, αS fibrils induced the formation of large Aβ42 oligomers while preventing these oligomeric species from further fibrillization, which was not previously observed with Aβ40. We also provide evidence that the Aβ42-αS interactions were mediated by various parts of Aβ42, such as the N-terminus, Aβ (22-35) and the C-terminus, depending on Aβ42 and αS conformers. Thus, the present study illustrates the nature of Aβ-αS interactions, both unique to Aβ42-αS and common to those observed with Aβ40-α.

2. MATERIALS AND METHODS

2.1. Reagents

Human β-amyloid (1-42) (Aβ42), synthesized via solid-phase chemistry, was purchased from ERI Amyloid Laboratory (Oxford, CT, USA). Aβ42, conjugated with HiLyte Fluor 488 at the N-terminus, was purchased from AnaSpec (Fremont, CA, USA). Alexa Fluor 647 NHS ester, for N-terminal αS labeling, was purchased from Thermo Fisher Scientific (Waltham, MA, USA). Aβ sequence-specific antibodies, 6E10, 4G8, and 12F4 were purchased from Biolegend (San Diego, CA, USA). The anti-Aβ(22-35) antibody was purchased from Sigma-Aldrich (St Louis, MO, USA). αS sequence-specific antibodies, F11, 5C2, 211 and D10, were purchased from Santa Cruz Biotechnology, Inc (Dallas, TX, USA) and Novus International (Saint Charles, MO, USA). Prepacked FPLC columns were purchased from GE Healthcare (Piscataway, NJ, USA), including HiTrap Q Sepharose HP column for anion exchange, HiPrep 16/60 Sephacryl S-100 HR column for size exclusion, and HiPrep 26/10 Sephadex column for desalting. All other chemical reagents were purchased from Fisher Scientific (Pittsburg, PA, USA) unless stated otherwise.

2.2. Preparation of monomeric Aβ42 and αS

Monomeric Aβ42 was prepared according to a well-established protocol [59, 60]. Briefly, lyophilized stock of Aβ42 was pretreated with hexafluoroisopropanol (HFIP) to dissociate any pre-existing aggregates. This was completed by addition of HFIP to the Aβ42 at a concentration of ~2 mg/mL and followed by incubation at room temperature for 2 hours. The Aβ solution was then aliquoted into separate tubes and vacuumed dried overnight to remove the volatile HFIP. The vacuum dried samples were stored at −80 °C until further use. To prepare monomeric Aβ42, a tube of the dried sample was reconstituted in 20 mM NaOH at a concentration of ~300 μM and incubated at room temperature for 20 minutes. The NaOH treatment promotes the disaggregation of pre-formed aggregates [61]. The samples were diluted with 220 μL of ice-cold dH2O, 65 μL of 30X phosphate buffer with azide (PBA; 1X PBA: 20 mM Na₂HPO₄/NaH₂PO₄, 0.02% NaN₃, pH 7.2) and 65 μL of 10X phosphate buffered saline with azide (PBSA; 1X PBSA: 20 mM Na₂HPO₄/NaH₂PO₄, 150 mM NaCl, 0.02% NaN₃, pH 7.2). The final concentration of the buffer is 80 mM Na₂HPO₄/NaH₂PO₄, 150 mM NaCl and 0.02% NaN₃. The Aβ42 solution was then filtered through a 0.45 μm syringe filter and the concentration was calculated by absorbance at 280 nm (A₂₈₀) with correction for scattering effects [62] on a Varian Cary 50 UV-Vis spectrophotometer (Agilent Technologies, Santa Clara, CA, USA).

To prepare αS monomers, the lyophilized αS was resuspended in PBSA. After ensuring that αS is fully dissolved, the solution was filtered through an EMD Millipore Amicon Ultra 0.5 mL 100 kDa MWCO filter (Millipore Sigma, Burlington, MA, USA). We used 100 kDa MWCO filters to prepare αS monomers, because, due to the natively unfolded nature, their hydrodynamic size is similar to that of ~60 kDa globular proteins [63]. The αS monomers were collected in the filtrate and kept on ice until further use.

2.3. Preparation of oligomeric Aβ42 and αS

Aβ oligomers were prepared by incubation of Aβ monomers. Briefly, the Aβ monomers were diluted to a concentration of 100 μM and incubated at 25 °C for 15 hours with no agitation (i.e., neither shaking nor stirring). Post incubation, the Aβ samples were transferred to siliconized microcentrifuge tubes and centrifuged at 16,900 xg for 5 minutes to remove the precipitates. The supernatant, containing the Aβ oligomers, was collected and Aβ42 concentration was estimated by A₂₈₀ with correction for scattering effects [62] in 8 M urea. Aβ42 oligomer samples were kept on ice prior to immediate use.

For αS oligomers preparation, lyophilized αS protein was reconstituted in PBSA and filtered through a 0.45 μm syringe filter. The αS concentration was measured using A₂₈₀ and adjusted to 350 μM (~5 mg/mL) prior to transferring to a glass vial (up to a max volume of 600 μL in each). Glass vials containing the αS solution were subjected to orbital shaking at 250 rpm in a 37 °C incubator for 6 hours. After the culmination of 6 hours, the αS solution was filtered using an Amicon Ultra 100 kDa MWCO filter. The retentate was washed 3 times with PBSA to remove any residual αS monomers and αS oligomers was collected in the retentate. Due to the relatively low fraction (~5 %) of αS oligomers being formed during preparation [64], the stock concentration of pre-formed αS oligomers was ≤ ~70 μM.

Aβ42 and αS concentrations used in this study are monomer-equivalent concentrations.

2.4. Preparation of fibrillar Aβ42 and αS

Fibrillar forms of Aβ42 and αS were produced by prolonged incubation of the Aβ42 monomer and αS monomer samples, which were prepared as described above. Aβ42 monomers were diluted to 222 μM and incubated in a glass vial at 37 °C with stirring for 2-3 weeks. After the culmination of 2-3 weeks, the Aβ42 samples were transferred to siliconized centrifuge tubes and pelleted by centrifugation at 16,900 xg for 5 minutes. Prior to use, the pellet was washed 3 times by resuspension and centrifugation, to remove soluble Aβ42 species (i.e., monomers and oligomers) from the supernatant. αS fibrils were formed and isolated in a similar fashion, but at a concentration of 350 μM. Concentrations of the Aβ42 and αS fibrils were estimated by back calculation of the concentrations of Aβ42 and αS, respectively, removed in the washes via UV-Vis spectroscopy at A₂₈₀, with correction for scattering effect [62]. Aβ42 and αS concentrations used in this study are monomer-equivalent concentrations.

2.5. Preparation of Aβ42 and αS mixtures

Monomers, oligomers, and fibrils of Aβ42 and αS were prepared as described above and mixed in PBSA to a final volume of 400 μL. The mixture samples were incubated in siliconized tubes at 37 °C under quiescent conditions for a 7-day incubation period. Aliquots were taken on specific days to analyze and characterize aggregation states of the samples.

For selective monitoring of Aβ42 and αS on SDS- and Native-PAGE gels, fluorescently labeled polypeptides were mixed with their unlabeled counterpart at a ratio of 1:1000 (labeled: unlabeled) for Aβ42 in all three forms, 1:1000 for αS monomers and fibrils, while αS oligomers were prepared at a 1:100 ratio. Fluorescently labeled mixtures were incubated while covered from light, to minimize photodegradation of the fluorophores during the 7-day incubation period.

2.6. Thioflavin T (ThT) fluorescence

During incubation, samples were aliquoted and subjected to measurements of Thioflavin T (ThT) fluorescence. Briefly, 10 μL of 0.1 mM ThT was added to 10 μL of the aliquots and 180 μL of PBSA to a final volume of 200 μL. ThT fluorescence was measured on a Photon Technology International QuantaMaster4 Fluorometer (Horiba, Kyoto, Japan) with excitation set at 440 nm and the emission wavelength monitored at 485 nm.

2.7. Circular dichroism (CD) spectroscopy

Secondary structure of Aβ42 and αS samples were analyzed with circular dichroism on a Jasco J-815 CD spectrometer (Jasco Inc., Easton, MA, USA). Immediately prior to CD measurements, the samples were diluted into PBSA, if necessary, loaded into a cuvette with a 1 mm pathlength, and the CD spectra recorded from 205-250 nm. The sample spectra were adjusted by subtracting the buffer background spectrum and converted to mean residue molar ellipticity (deg·cm²·dmol^-1).

2.8. SDS-PAGE and Native-PAGE in-gel fluorescence

A combination of SDS-PAGE and Native-PAGE was used to analyze the aggregation states of the samples. The gels ran on the XCell Surelock Mini-Cell gel tank (Thermo Fisher Scientific), with different running buffers for SDS-PAGE and Native-PAGE.

For SDS-PAGE, total and soluble fractions were prepared and analyzed separately. To prepare the soluble fraction, 50 μL of the sample was centrifuged at 16,900 xg for 5 minutes and the supernatant was used as the soluble fraction. Three μL of Fluorescent Compatible Sample Loading Buffer (Thermo Fisher Scientific) was added to 9 μL of each sample and heated at 95 °C for 15 minutes. Samples were loaded onto a NuPAGE 4-12% Bis-Tris Gel and run at 200 V with NuPAGE MES SDS Running Buffer for 25 minutes (Thermo Fisher Scientific).

For Native-PAGE, 9 μL of the samples were loaded into each well of the Native-PAGE 4-16% Bis-Tris Gel (Thermo Fisher Scientific). Electrophoresis was performed at 150 V with Novex Tris-Glycine Native running buffer for 90 minutes.

Samples containing HiLyte Fluor 488 labeled Aβ42 (green) and Alexa Fluor 647 labeled αS (red), were utilized for fluorescence imaging of SDS- and Native-PAGE gels. The gels were imaged on an Amersham Typhoon RGB (Cytiva, Marlborough, MA, USA) located at the Genomics Core at NYU Center for Genomics and Systems Biology.

2.9. Transmission electron microscopy (TEM)

Five μL of the samples were pipetted onto 400 mesh copper grids and then negatively stained with 1% uranyl acetate in dH2O solution. The grids were imaged on a FEI Talos L120C Transmission Electron Microscope (FEI Corp, Hillsboro, OR, USA) with a 4k x 4k GATAN OneView camera. The Transmission Electron Microscope is located at the Microscopy Laboratory at NYU Langone Medical Center.

2.10. Competitive binding dot blot assay

A competitive binding dot blot assay was utilized to determine the regions of Aβ42 to which αS monomers, oligomers and fibrils bind. Mixtures of Aβ and αS were incubated at a 1:10 ratio (35 μM Aβ42 : 350 μM αS) to ensure saturation of the binding sites on Aβ with αS. A total of 1 μg of Aβ was dotted onto the nitrocellulose membrane, from each samples, and the membrane was air-dried for 15 minutes, followed by blocking, washing and incubation with primary antibodies and alkaline phosphatase-conjugated secondary antibodies, and chemiluminescent development according to the manufacturer’s protocols. The chemiluminescent membrane was imaged on an Amersham Imager 680 (Cytiva, Marlborough, MA, USA). The Aβ42 sequence specific primary antibodies used in this assay included 6E10, 4G8, anti-Aβ (22-35) and 12F4.

2.11. Dynamic light scattering

A Zetasizer Nano ZS 90 (Malvern Panalytical, Malvern, United Kingdom) was utilized to examine size distribution of the samples using CONTIN algorithm. Distribution results reported were the average of three measurements, with each measurement captured over the course of 30 seconds.

Other experimental details including αS expression and purification and fluorescent labeling of αS can be found in Supplementary Material.

3. RESULTS

In our experiments to elucidate Aβ42-αS interactions in amyloid aggregation, we co-incubated Aβ42 with αS in excess, unless otherwise stated, because αS is more abundant than Aβ42 in physiological conditions [65, 66]. The concentration of Aβ42 was set at 35 μM so that its aggregation and impacts of αS on Aβ42 can be readily determined within a reasonable timeframe under our in vitro experimental condition (37 °C without any shaking or agitation). Moreover, though the Aβ42 and αS concentrations used in this study were higher than physiological concentrations [67-70], there exist various mechanisms in vivo, where Aβ42 and αS concentrations can be elevated locally [68, 71-75]. Thus, our in vitro study described herein still provide important insight into nature of Aβ42-αS interactions and their consequences.

In our study herein, Aβ42-αS interactions were examined using a comprehensive array of experimental tools: fluorescence of Thioflavin T (ThT), a dye recognizing β-sheet rich amyloid aggregates [76], transmission electron microscopy (TEM) for aggregate morphology and SDS- and Native-PAGE for distributions among monomers, oligomers and fibrils in samples. For selective monitoring of Aβ42 and αS on SDS- and Native-PAGE, samples containing HiLyte Fluor 488-labeled Aβ42 and/or Alexa Fluor 647-labeled αS were used. Note that these N-terminal labelings did not affect aggregation properties of Aβ or αS, as judged by ThT fluorescence, size exclusion chromatography, dynamic light scattering and electrophoretic mobility [57, 64, 77-80] (also see Fig. S1).

Control samples of Aβ42 monomers, Aβ42 oligomers, Aβ42 fibrils, αS monomers, αS oligomers and αS fibrils were prepared and characterized during incubation at 37 °C under a quiescent condition for 7 days. In short, Aβ42 monomers and Aβ42 oligomers aggregated into insoluble fibrillar aggregates and Aβ42 fibrils remained fibrillar, whereas there was no significant change in the αS controls during the 7-day incubation (Supplementary Text, Fig. S2-S4 and Table 1).

Table 1.

Summary of the major findings.

Sample		The major aggregate species observed after 7-day incubation
Aβ42 M	Control	Insoluble fibrils
	+ αS M	Soluble oligomers and insoluble fibrils
	+ αS O	Soluble oligomers
	+ αS F	Large Aβ42 oligomers and insoluble fibrils
Aβ42 O	Control	Insoluble fibrils
	+ αS M	Soluble oligomers
	+ αS O	Soluble oligomers
	+ αS F	Large Aβ42 oligomers and insoluble fibrils
Aβ42 F	Control	Insoluble fibrils
	+ αS M	Insoluble fibrils
	+ αS O	Insoluble fibrils
	+ αS F	Insoluble fibrils

M, O and F represent monomers, oligomers, and fibrils, respectively.

3.1. Aβ42 monomers and αS monomers

The impact of αS monomers (350 μM) on the aggregation of Aβ42 monomers (35 μM) was examined by co-incubation over the course of 7 days. While the ThT fluorescence profile of samples containing Aβ42 and αS monomers largely aligns with the Aβ42 monomer control sample (Fig. 1A), TEM images of the mixture on Day 7 showed the presence of aggregates that were mostly protofibrillar, often clumped together (Fig. 1B), along with some short fibrillar species (Fig. S5). Consistent with this finding, the addition of αS monomers interfered with conversion of soluble Aβ42 monomers to insoluble aggregates, though not completely (Fig. 1E). A similar effect was also found with αS monomers at 70 μM (Fig. S6). This result is in contrast with the Aβ42 monomers control, which was fully lost to insoluble fibrils (Fig. S2C). Note that aggregation of Aβ40 monomers at 75 μM into insoluble species was completely prevented by αS monomers at 350 μM under the otherwise same 7-day incubation condition in our previous study [57]. The implication is that αS monomer’s aggregation inhibitory effects appeared stronger on Aβ40 than Aβ42 monomers. The dominant fraction of Aβ42 soluble species after the 7-day incubation with αS monomers was oligomeric, as indicated by the smeared bands located in the upper portion of the Native-PAGE gel (Fig. 1F). Slight overlaps of oligomeric bands from each fluorophore were noticeable in the upper portion of the Native-PAGE gel (Fig. 1F), which may result from binding (or co-assembly) between Aβ42 and αS. The oligomers present in the mixture on Day 7 consisted of two major populations with ~70 and ~190 nm in hydrodynamic diameter (d_h) (Fig. 1G), which are similar to the size range for the freshly prepared Aβ42 oligomer control (Fig. S2G). During the 7-day incubation, αS monomers in the mixtures remained soluble and monomeric (Fig. 1E-F), as was observed with the αS monomer controls (Fig. S4C-D). Consistent with this finding, a population with d_h of ~7 nm was detected in the mixture (Fig. 1G), which is in a similar size range of freshly prepared αS monomers (Fig. S4E). Despite the non-negligible, yet modest, aggregation inhibitory effects of αS monomers on Aβ42 monomers, time-course ThT fluorescence looked similar whether or not αS monomers were added to Aβ42 monomer samples. This is presumably because (1) ThT-positive Aβ42 fibrillar aggregates were still produced in the mixture samples due to the incomplete inhibition of Aβ42 monomer aggregation by αS monomers and (2) Aβ42 oligomers generated in the mixture samples can also be ThT-positive [60], though not necessarily at the same level as Aβ42 oligomer controls (Fig. S2A) if these oligomers differ in local structure of ThT binding sites [76]. Taken together, αS monomers inhibited the fibrillization of Aβ42 monomers, though not as fully as Aβ40 monomers, while promoting conversion of these Aβ42 species into ThT-positive oligomeric assemblies.

Figure 1.

Characterization of Aβ42 monomers mixed with αS monomers (M), αS oligomers (O) and αS fibrils (F) incubated for 7 days at 37 °C, as examined by (A) ThT fluorescence, (B-D) TEM, in-gel fluorescence imaging of (E) SDS-PAGE and (F) Native-PAGE, and (G-H) size distribution as determined by CONTIN analysis of dynamic light scattering. The data on Aβ42 monomers mixed with αS monomers (filled red diamonds), αS oligomers (filled blue squares) and αS fibrils (filled grey circles) are shown along with those on αS monomers alone (empty red diamonds), αS oligomers alone (empty blue squares) and αS fibrils alone (empty gray circles) taken from Fig. S4A. The data on Aβ42 monomers alone (black exes) taken from Fig. S2A are also shown for comparison. In (A), error bars: 1 standard deviation of triplicates. In (B-D), representative TEM images of mixtures on Day 7 containing (B) Aβ42 M and αS M, (C) Aβ42 M and αS O, and (D) Aβ42 M and αS F are shown with scale bars of 200 nm. The Aβ42 M control on Day 7 (Fig. S2B) is shown next to Fig. 1D for comparison. In (E-F), Aβ42 and αS samples contained HiLyte Fluor 488-labeled Aβ42 (green) and Alexa Fluor 647-labeled αS (red), respectively. Each panel was taken from a bigger gel image and reassembled for better presentation. The SDS- and Native-PAGE gel images of Aβ42 M control on Day 7 (Figs. S2C-D) are shown for comparison. In (E), T: total fraction and S: soluble fraction. In (F), the image to the right provides a brighter image of the upper portion (enclosed in the blue box) of the Native-PAGE gel, which was obtained by imaging only the upper portion gel; to renormalize the brightness to show the oligomer bands. Fluorescent intensity of the monomeric bands often overwhelmed the presence of oligomer bands, which may possess a lower percentage of the fluorophores. G: green channel and R: red channel. (G-H) Size distribution for the mixtures of (G) Aβ42 M and αS M and (H) Aβ42 M and αS O on Day 7.

3.2. Aβ42 monomers and αS oligomers

The addition of αS oligomers (17 μM) to Aβ42 monomers (35 μM) had a similar modulatory effect on Aβ42 aggregation, as αS monomers (350 μM) had, although the concentration of αS oligomers was ~20-times lower. Co-incubation of Aβ42 monomers with αS oligomers for 7 days resulted in the formation of ThT-positive, protofibrillar and globular aggregates (Figs. 1A and 1C). The aggregates in the mixtures were deemed stable in the late stage of the incubation, as there was no significant ThT fluorescence changes after Day 1 (Fig. 1A). During the 7-day incubation, αS oligomers inhibited the formation of insoluble aggregates from Aβ42 monomers (Fig. 1E). Instead, there was a strong overlap of oligomeric bands from each fluorophore in the upper portion of the Native-PAGE gel (Fig. 1F), which may result from binding (or co-assembly) between Aβ42 and αS in the oligomeric state. The interaction between Aβ42 monomers and αS oligomers produced an additional fraction of SDS-resistant Aβ42 oligomers, indicated by multiple bands above the monomeric band on the SDS-PAGE gel (Fig. 1E; compared to the Aβ42 oligomer control in Figs. S1 and S2C). On DLS, two major populations with d_h of ~20 and ~65 nm were detected for the mixture on Day 7 (Fig. 1H): the smaller d_h is likely indicative of the size of αS oligomers that remained soluble and oligomeric in the presence of Aβ42 monomers (Figs. 1E-F; compared to d_h of the αS oligomer controls in Fig. S4G), whereas the larger d_h is presumably assigned to the size of the Aβ42 oligomers possibly bound to or co-assembled with αS. Overall, αS oligomers induced the formation of stable oligomeric assemblies by inhibiting the fibrillization of Aβ42 monomers.

3.3. Aβ42 monomers and αS fibrils

To comprehensively examine the αS conformer-dependent effect on the aggregation of Aβ42 monomers, we also co-incubated these Aβ42 species (35 μM) with αS fibrils (350 μM). The mixture remained relatively steady over the duration of the experiment, with ThT fluorescence at a level similar to the αS fibril controls (Fig. 1A) and in fibrillar morphology (Fig. 1D). Consistent with this finding, no additional αS fibril dissociation was observed (Figs. 1E-F), compared to the αS fibril controls (Figs. S4C-D). Interestingly, SDS-PAGE indicates that a fraction of Aβ42 soluble species remained after Day 7 (Fig. 1E), though no bands were visible on the Native-PAGE gel (Fig. 1F). The implication is that αS fibrils induced the formation of large, soluble Aβ42 oligomers, which were unable to enter the Native-PAGE gel. No such effect of αS fibrils were reported with Aβ40 monomers under an otherwise similar condition [57].

3.4. Aβ42 oligomers and αS monomers

The study was then extended to pre-formed Aβ42 oligomers (35 μM) with αS monomers (350 μM) (Fig. 2). ThT fluorescence of Aβ42 oligomers in the presence of αS monomers was largely similar to the Aβ42 oligomer control for 7 days (Fig. 2A). TEM showed the presence of protofibrils (Fig. 2B), often surrounded by clumps of globular aggregates in this mixture on Day 7 (Fig. S7). Despite a similar ThT increase to the Aβ42 oligomer control (Fig. 2A), which aggregated into insoluble species during the incubation (Fig. S2C), Aβ42 oligomers remained soluble oligomeric when co-incubated with αS monomers (Fig. 2E-F). Consistent with this finding, a DLS analysis of the mixture after the 7-day incubation detected populations with d_h of ~ 7 nm as well as ~60 and ~250 nm (Fig. 2G), which correspond to the sizes of freshly prepared αS monomers and Aβ42 oligomers, respectively (Figs. S2G and S4E). Slight overlaps of oligomeric bands in the upper portion of the Native-PAGE gel was observed on Day 7 (Fig. 2F), which may result from binding (or co-assembly) between Aβ42 and αS. Despite the aggregation inhibitory effects of αS monomers on Aβ42 oligomers, ThT fluorescence changed over time at similar rates between the mixture samples and Aβ42 oligomer controls. The implication is that (1) soluble Aβ42 oligomers grew at similar rates in both presence and absence of αS monomers and (2) Aβ42 oligomers remaining soluble in the presence of αS monomers shared molecular similarity with the relatively low ThT positive Aβ42 aggregates produced from incubation of Aβ42 oligomer samples (Fig. S2A).

Figure 2.

Characterization of Aβ42 oligomers mixed with αS monomers (M), αS oligomers (O) and αS fibrils (F) incubated for 7 days at 37 °C, as examined by (A) ThT fluorescence, (B-D) TEM, in-gel fluorescence imaging of (E) SDS-PAGE and (F) Native-PAGE, and (G-H) size distribution as determined by CONTIN analysis of dynamic light scattering. The data on Aβ42 oligomers mixed with αS monomers (filled red diamonds), αS oligomers (filled blue squares) and αS fibrils (filled grey circles) are shown along with those on αS monomers alone (empty red diamonds), αS oligomers alone (empty blue squares) and αS fibrils alone (empty gray circles) taken from Fig. S4A. The data on Aβ42 oligomers alone (black exes) taken from Fig. S2A are also shown for comparison. In (A), error bars: 1 standard deviation of triplicates. In (B-D), representative TEM images of mixtures on Day 7 containing (B) Aβ42 O and αS M, (C) Aβ42 O and αS O, and (D) Aβ42 O and αS F are shown with scale bars of 200 nm. The Aβ42 O controls on Day 7 (Fig. S2B) is shown next to Fig. 2D for comparison. In (E-F), Aβ42 and αS samples contained HiLyte Fluor 488-labeled Aβ42 (green) and Alexa Fluor 647-labeled αS (red), respectively. Each panel was taken from a bigger gel image and reassembled for better presentation. The SDS- and Native-PAGE gel images of Aβ42 O control on Day 7 (Figs. S2C-D) are shown for comparison. In (E), T: total fraction and S: soluble fraction. In (F), the image to the right provides a brighter image of the upper portion (enclosed in the blue box) of the Native-PAGE gel, which was obtained by imaging only the upper portion gel; to renormalize the brightness to show the oligomer bands. Fluorescent intensity of the monomeric bands often overwhelmed the presence of oligomer bands, which may possess a lower percentage of the fluorophores. G: green channel and R: red channel. (G-H) Size distribution for the mixtures of (G) Aβ42 O and αS M and (H) Aβ42 O and αS O on Day 7.

3.5. Aβ42 oligomers and αS oligomers

Co-incubation of Aβ42 oligomers (35 μM) and αS oligomers (17 μM) generated a ThT fluorescence signal that remained relatively constant during the 7-day incubation (Fig. 2A). A group of protofibrils surrounded by globular oligomers was detected in the mixture sample on Day 7 (Fig. 2C). In-gel fluorescence imaging confirmed the stabilization of Aβ42 oligomers by αS oligomers, which was likely by binding or co-assembly between the two species (Fig. 2E-F). The formation of additional SDS-resistant Aβ42 oligomers upon the 7-day co-incubation with αS oligomers was also noticeable (Fig. 2E), as described for the mixture of Aβ42 monomers and αS oligomers (Fig. 1E). When evaluated by DLS, the mixture samples contained the three major populations on Day 7 (Fig. 2H): the d_h values (~65 and ~250 nm) of the two smaller populations fall within the size range of freshly prepared Aβ42 oligomers (Fig. S2G), suggesting that the largest population with ~800 nm in d_h is likely indicative of oligomeric species resulting from direct binding or co-assembly between Aβ42 and αS oligomers.

3.6. Aβ42 oligomers and αS fibrils

During the 7-day incubation, ThT fluorescence of the mixture containing Aβ42 oligomers (35 μM) and αS fibrils (350 μM) remained mostly unchanged at an intensity level similar to the αS fibril control (Fig. 2A). TEM imaging of this mixture on Day 7 (Fig. 2D) detected fibrils that were morphologically similar to the αS fibril controls (Fig. S4B). As was observed with the mixture containing Aβ42 monomers and αS fibrils (Fig. 1E), a fraction of Aβ42 oligomers remained soluble after the 7-day incubation (Fig. 2E) without being readily able to enter the Native-PAGE gel in the presence of αS fibrils (Fig. 2F). Thus, αS fibrils promoted the formation of large, soluble Aβ42 oligomers. αS fibrils in the presence of Aβ42 oligomers were similar to the αS fibril controls on SDS- and Native-PAGE (Fig. 2E-F vs. Fig. S4C-D), indicating the lack of any notable impact of Aβ42 oligomers on αS fibrils.

3.7. Aβ42 fibrils and αS in varying aggregation states

The mixture containing Aβ42 fibrils with each of the three αS conformers remained ThT positive at a value similar to either the Aβ42 fibril control (for co-incubation with αS monomers and oligomers) or the αS fibril control (for co-incubation with αS fibrils) (Fig. 3A). TEM imaging showed the existence of fibrils in all three combinations on Day 7, though morphologically different (Fig. 3B-D). Aβ42 fibril-like aggregates, surrounded by other aggregate species, were observed in the mixtures of Aβ42 fibrils with αS monomers and Aβ42 fibrils with αS oligomers (Fig. 3B-C), whereas αS fibril-like aggregates found in the mixture of both Aβ42 and αS fibrils (Fig. 3D). Aβ42 fibrils were largely insoluble throughout the duration of the co-incubations (Fig. 3E-F). No notable impact of Aβ42 fibrils on the three αS conformers was observed on SDS- and Native-PAGE (Fig. 3E-F), when compared to the αS controls (Fig. S4C-D).

Figure 3.

Characterization of Aβ42 fibrils mixed with αS monomers (M), αS oligomers (O) and αS fibrils (F) incubated for 7 days at 37 °C, as examined by (A) ThT fluorescence, (B-D) TEM, and in-gel fluorescence imaging of (E) SDS-PAGE and (F) Native-PAGE. The data on Aβ42 fibrils mixed with αS monomers (filled red diamonds), αS oligomers (filled blue squares) and αS fibrils (filled grey circles) are shown along with those on αS monomers alone (empty red diamonds), αS oligomers alone (empty blue squares) and αS fibrils alone (empty gray circles) taken from Fig. S4A. The data on Aβ42 fibrils alone (black exes) taken from Fig. S2A are also shown for comparison. In (A), error bars: 1 standard deviation of triplicates. In (B-D), representative TEM images of mixtures on Day 7 containing (B) Aβ42 F and αS M, (C) Aβ42 F and αS O, and (D) Aβ42 F and αS F are shown with scale bars of 200 nm. The Aβ42 F control on Day 7 (Fig. S2B) is shown next to Fig. 3D for comparison. In (E-F), Aβ42 and αS samples contained HiLyte Fluor 488-labeled Aβ42 (green) and Alexa Fluor 647-labeled αS (red), respectively. Each panel was taken from a bigger gel image and reassembled for better presentation. The SDS- and Native-PAGE gel images of Aβ42 F control on Day 7 (Figs. S2C-D) are shown for comparison. In (E), T: total fraction and S: soluble fraction. Aβ42 fibrils mixed with αS monomers did not readily enter the SDS-PAGE gel. In (F), the image to the right provides a brighter image of the upper portion (enclosed in the blue box) of the Native-PAGE gel, which was obtained by imaging only the upper portion gel; to renormalize the brightness to show the oligomer bands. Fluorescent intensity of the monomeric bands often overwhelmed the presence of oligomer bands, which may possess a lower percentage of the fluorophores. G: green channel and R: red channel.

Overall, (1) soluble αS species (i.e., monomers and oligomers) enhanced Aβ42 oligomerization and stabilized Aβ42 oligomers, while preventing Aβ42 fibrillization, and (2) αS fibrils promote the formation of large soluble Aβ42 oligomers. The major findings on the mixtures of Aβ42 and αS are summarized in Table 1.

3.8. Competitive binding dot blot assay

To reveal the potential Aβ42 regions with which αS interacts, a competitive binding dot blot assay with Aβ42 sequence-specific antibodies was performed. To ensure that most of Aβ42 is bound to αS, an excess of αS was added (35 μM Aβ42 : 350 μM αS). If αS binds to a specific region of the Aβ sequence, it would hinder the corresponding antibody from binding, thus producing a reduced signal. The Aβ sequence-specific antibodies utilized included 6E10 (for Aβ 1-16), 4G8 (for Aβ 17-22), Anti-Aβ (22-35) (for Aβ 22-35) and 12F4 (for Aβ 36-42), covering the entire Aβ42 sequence. These antibodies bound to the three Aβ42 conformers (i.e., monomers, oligomers and fibrils), but not to αS, as expected (data not shown). No similar study was performed with αS sequence-specific antibodies to determine Aβ42 binding sites on αS due to these antibodies’ relatively long epitopes or incomplete coverage of αS.

From examination of the membranes, the most notable weakening of antibody signals occurred with 12F4 (Fig. 4A-D), indicating that interactions with αS is primarily mediated by Aβ42’s C-terminus. Strong interference by soluble αS species (i.e., monomers or oligomers) on 12F4’s binding to Aβ42 monomer (Fig. 4D) is consistent with the observation of direct binding or co-assembly between these Aβ42 and αS species, as described above. In addition to the Aβ42 C-terminus, the Aβ (22-35) fragment also seemed to mediate, though relatively weakly, interactions between Aβ42 and αS monomers (Fig. 4C). Similarly, interactions of Aβ42 oligomers via their C-termini with αS monomers was evident (Fig. 4D). In contrast, despite the large overlap of fluorescence between Aβ42 and αS oligomers on the Native PAGE gels (Fig. 2F), no strong inhibition of antibody binding was detected for this Aβ42-αS mixture (Fig. 4). This might be due to a large fraction of Aβ42 chains in their oligomeric state available for antibody binding, whereas only its small fraction being directly involved in binding to αS oligomers. Alternatively, the binding site on Aβ42 oligomers for interactions with αS oligomers might consist of short Aβ42 segments originated from multiple Aβ42 chains.

Figure 4.

Competitive binding dot blot assay using Aβ42 sequence-specific antibodies, (A) 6E10, (B) 4G8, (C) Anti-Aβ (22-35), and (D) 12F4. Aβ42 monomers (M, top panels), Aβ42 oligomers (O, middle panels) and Aβ42 fibrils (F, bottom panels) were mixed with αS M (left column), αS O (middle column) and αS F (right column) and subsequently dotted on the nitrocellulose membrane at 1 μg of Aβ42 each. Chemiluminescence intensities of dot blot signals were quantified by ImageQuant TL 8.1 and normalized to the dot blot signals of the Aβ42 controls. Error bars: 1 SD (w = 3).

The competitive binding dot blot assay also indicates that all forms of Aβ42 binds to αS fibrils via the Aβ42 C-terminus as the primary binding site (Fig. 4). Though not as strongly as the Aβ42 C-terminus, Aβ42 N-terminus and the Aβ (22-35) fragment also seemed involved in binding between αS fibrils and different Aβ42 conformers (Fig. 4). The interactions are likely to be responsible for the observed inhibition by αS fibrils of soluble Aβ42 species’ conversion into insoluble aggregates (Figs. 1E >and 2E). On the other end, the binding between Aβ42 and αS fibrils evident in the dot blot assay (as exemplified in Fig. 4D) resulted in no significant changes in terms of ThT fluorescence, TEM and SDS- and Native-PAGE (Fig. 3), compared to the Aβ42 or αS fibril controls. The lack of any change in the mixture is probably because (1) αS fibrils was in excess relative to Aβ42 fibrils, dominating ThT fluorescence signals and representing the major fraction on TEM and (2) both Aβ42 and αS fibrils remained largely insoluble in the mixture, unable to enter SDS- or Native-PAGE gels.

4. DISCUSSION

In this study, we showed that the three αS conformers (i.e., monomers, oligomers and fibrils) inhibited the fibrillization of Aβ42 monomers and oligomers (Fig. 5). Among the various Aβ42-αS combinations tested in this study, the most drastic change was observed when Aβ42 and αS oligomers were co-incubated. The addition of αS oligomers to Aβ42 oligomers maintained ThT fluorescence of the mixture at a level similar to the freshly prepared Aβ42 oligomer control (Fig. 2A). The 7-day incubation of the mixture produced oligomeric species, whose d_h value was beyond the size ranges of the Aβ42 and αS oligomer controls (Fig. 2H). Moreover, between αS monomers and αS oligomers, the latter was more effective than the former at promoting the formation of soluble oligomers from Aβ42 monomers: non-negligible oligomerization occurred from Aβ42 monomers immediately after mixing with αS oligomers at 17 μM, whereas no such effect was observed with αS monomers at 350 μM on Day 0 (Fig. 1F). Additionally, ThT fluorescence of the mixtures containing soluble Aβ42 species (i.e., monomers and oligomers) became stabilized in the presence of αS oligomers at 17 μM but not αS monomers at 350 μM (Fig. 1A and Fig. 2A). The stronger effects of αS oligomers may be due in part to multivalent interactions [57, 81], where multiple αS chains in close proximity are more effective than individual αS monomers.

Figure 5.

Schematic summarizing the interactions between Aβ42 and αS. Aβ42 monomers self-assembled into Aβ42 oligomers, which then further aggregated to form Aβ42 fibrils. Addition of αS soluble species (i.e., monomers and oligomers) promoted oligomerization of Aβ42 monomers and stabilizes pre-formed Aβ42 oligomers possibly via co-assembly or direct binding. αS fibrils promoted the formation of large oligomeric aggregates from Aβ42 monomers and oligomers. Whether these large oligomers contain αS is currently unknown. Fibrillar forms of Aβ42 were incorporated into αS fibrils. M, O and F represent monomers, oligomers and fibrils, respectively.

Our findings described in the current study with Aβ42 and the previous study with Aβ40 illustrates both similarities and differences in the consequences of interactions with αS. For example, αS fibrils induced the formation of large Aβ42 oligomers from Aβ42 soluble species (e.g., via heterogeneous nucleation of Aβ42 aggregates [55]), while interfering, though not fully, with fibrillization of these Aβ42 species. In contrast, no such effect of αS fibrils was observed on Aβ40 soluble species [57]. Interestingly, while binding of αS fibrils to the Aβ C-terminus was detected for soluble species of both Aβ42 (in the current study) and Aβ40 [57], the Aβ N-terminus and Aβ (22-35) seemed involved, though not as strongly as the Aβ C-terminus, only in Aβ42-αS, but not Aβ40-αS interactions. Notably, Aβ40 and Aβ42 differ in outcomes of familial mutations at the N-terminus (e.g., A2 and D7) in terms of oligomerization and fibrillization (see [82] for review). Collectively, these results suggest that binding of Aβ to αS may occur at different locations and produce different consequences, depending on the presence and absence of the last two C-terminal amino acids of Aβ42 (Ile-Ala). The C-terminal end of Aβ42 is more structured and more critical to its aggregation than that of Aβ40 [83, 84]. Therefore, αS fibril binding to the C-terminus of soluble Aβ42 species may disrupt their fibrilization more effectively than Aβ40.

Aβ42 and Aβ40 also shared similarities in the outcome of interactions with αS. Soluble αS species promoted oligomerization of Aβ monomers and stabilized pre-formed Aβ oligomers for both Aβ42 (this study) and Aβ40 [57], possibly by co-assembly and primarily mediated on the Aβ C-terminus. This similarity is noteworthy as Aβ40 and Aβ42 are known to follow distinct aggregation pathways to form soluble oligomers from monomers [85]. Familial mutations promoting αS oligomerization are found on the αS N-terminus (e.g., A30, E46 and A53) [16], implying that binding of Aβ to this location of αS might be responsible for the formation of co-assembled oligomers, which has yet to be validated. As was the case with Aβ40 fibrils [57], the competitive binding dot blot assay showed no strong binding between Aβ42 fibrils and αS soluble species (Fig. 4), though seeding αS aggregation by Aβ42 fibrils was previously reported [50]. This is presumably because the seeding may not require strong binding between Aβ42 and αS, or αS soluble species bind to surface of Aβ42 fibrils across rather than parallel to Aβ42 chains. The involvement of the Aβ N-terminus, Aβ (22-35) and the Aβ C-terminus for interaction between αS fibrils and Aβ42 soluble species, but not between Aβ42 fibrils and soluble αS species, may be related to differences in fibril structures (in terms of the number of β-strands and the U-turn domain) [86] and the seeding mechanisms between the two combinations [50].

The interactions of Aβ42 and αS has significant biological implications. The Aβ42 isoform is classified as the most toxic Aβ isoform [1-3, 58], and stabilization of Aβ42 oligomers may result in more neuronal cell death. The formation of large soluble oligomers by Aβ42, but not by Aβ40, in the presence of αS fibrils may represents an Aβ42-specific toxic mechanism. While it remains unknown if oligomeric αS can be released from αS fibrils during their interactions with soluble Aβ42, this release mechanism may provide additional toxic effects [87]. Note that direct binding or co-assembly between Aβ42 and αS may serve as a route for escaping from their cellular clearance by proteolysis [88], enhancing their biological effects. Further biological studies will be necessary in determining the biological consequences of Aβ42-αS interactions.

5. CONCLUSIONS

In conclusion, our present study demonstrates that (1) soluble αS monomers and oligomers promoted oligomerization and stabilization of soluble Aβ42 and (2) αS fibrils hindered soluble Aβ42 species from converting into insoluble aggregates by the formation of large oligomers. Interactions between Aβ42 and αS were facilitated primarily through the Aβ42 C-terminus, along with the Aβ42 N-terminus and Aβ (22-35), depending on Aβ42 and αS conformers. Overall, the research in this study demonstrates aggregation state-specific interactions between Aβ42 and αS and presents (1) additional insight into the synergistic and biological implications of the two polypeptides and (2) similarities to and differences from Aβ40-αS interactions.

ABBREVIATIONS

AD

Alzheimer’s disease

PD

Parkinson’s disease

Aβ

β-amyloid

αS

α-synuclein

APP

amyloid precursor protein

LB

Lewy bodies

NAC

non-amyloid-β component

DLB

dementia with Lewy bodies

PDD

PD with dementia

Tg

transgenic

SDS

sodium dodecyl sulfate

HFIP

hexafluoroisopropanol

PBA

phosphate buffer with azide

PBSA

phosphate buffered saline with azide

A₂₈₀

absorbance at 280 nm

ThT

Thioflavin T

CD

circular dichroism

PAGE

polyacrylamide gel electrophoresis

TEM

transmission electron microscopy

M

monomer

O

oligomer

F

fibril