Ultrasonic detection of $\alpha$-synuclein amyloid seeds from a highly crowded environment

Abstract

Detecting $\alpha$-synuclein ($\alpha$-Syn) amyloid seeds in biological fluids is a promising approach for the early diagnosis of Parkinson’s disease. However, detecting subtle amounts of seeds in highly crowded environments remains challenging. Ultrasonication can enhance seed detection by efficiently fragmenting fibrils, but its effects in crowded environments have not been fully explored. In this study, we apply ultrasonication to detect $\alpha$-Syn seeds in a highly crowded milieu and investigate its effects on seed detection. Our results show that ultrasonication enables rapid detection of $\alpha$-Syn seeds with a detection limit of 10 pg/mL, even in the presence of 40 mg/mL serum albumin. Intriguingly, the amount of fibril formed depends on the initial seed concentration in a crowded environment only under ultrasonication. To understand this phenomenon, we theoretically analyze the kinetics of seed-dependent amyloid formation. The results suggest that ultrasonic cavitation induces the formation of a dead-end complex between serum albumin and $\alpha$-Syn monomers, which can reduce false positives by suppressing seed-independent amyloid formation. These findings demonstrate ultrasonication as a powerful tool for the sensitive detection of $\alpha$-Syn seed in clinical diagnostics.

Significance

To eradicate amyloidosis, represented by Alzheimer’s and Parkinson’s disease, a methodology for early-stage diagnosis is vital. In this study, we explore the potential of ultrasonication to protein solution as a tool for the detection of amyloid seeds of $\alpha$-synuclein, a promising biomarker for the early diagnosis of Parkinson’s disease. The results show that ultrasonication can detect ultratrace amyloid seeds (10 pg/mL), even in a highly crowded milieu, which cannot be detected by a conventional method. We discuss the ultrasonic effects on seed detection from a biophysical viewpoint. The results indicate that ultrasonic cavitation plays a key role in the sensitive detection of seeds from a crowded sample. These results highlight the potential of ultrasonication as a methodology for early-stage diagnosis of amyloidosis.

Introduction

Amyloid fibrils are insoluble aggregates of denatured proteins with an ordered structure, implicated in various neurodegenerative diseases, such as Alzheimer’s disease and Parkinson’s disease (PD) (1,2). In these diseases, amyloid fibrils formed in vivo damage neuronal cells, subsequently leading to cognitive and/or motor malfunction. Given the significant challenges in treating these diseases after the onset of clinical symptoms, early diagnosis before symptoms appear is crucial to eradicating these diseases (3). Detecting short fragments of amyloid fibrils, commonly termed seeds, in cerebrospinal fluids and serum has emerged as a promising strategy for early diagnosis (4,5,6). However, the extremely low concentration of amyloid seeds in the biological fluids poses a substantial challenge for reliable detection.

Real-time quaking-induced conversion (RT-QuIC) is one of the most promising techniques for detecting amyloid seeds in biological samples. The effectiveness has especially been proven for distinguishing patients with PD from healthy individuals based on the seed detection of $\alpha$-synuclein ($\alpha$-Syn) (5,6,7,8), a protein responsible for PD (9,10). In the typical RT-QuIC assay, biological fluids are mixed with a solution of precursor monomers and exposed to shaking agitation while monitoring the formation kinetics of amyloid fibrils. If amyloid seeds are present in the sample, the fibril formation kinetics is accelerated, as the seeds bypass the rate-limiting nucleation step by acting as a template of amyloid formation (11). The resultant difference in fibril formation kinetics enables the distinction between patient and healthy samples (5,6). This assay is called the $\alpha$-Syn seed amplification assay (SAA). Several clinical reports have shown the potential of SAA for early diagnosis of PD (5,6,12). One of the clinical reports that successfully distinguished patients with PD from healthy individuals using SAA required an assay duration of more than 150 h (5), which limits the clinical utility of the method. This highlights the need for analytical techniques that can shorten the assay time for seed detection. Several studies have addressed this by optimizing the conditions of the solution under shaking agitation (13,14).

Ultrasonication has emerged as a powerful method for amplifying amyloid seeds and reducing the time frame required for seed detection (15,16,17,18,19). Ultrasonication induces the fragmentation of mature amyloid fibrils by the effects of ultrasonic cavitation (16,20). The ultrasonic fragmentation of fibrils increases the number of active fibril termini where the fibrils grow by attaching monomers, resulting in the rapid amplification of amyloid fibrils. Previous studies have demonstrated that ultrasonication improves detection sensitivity and shortens the detection time frame compared to conventional shaking methods (16,17), underscoring its robustness as a tool for amplifying amyloid seeds. In previous studies on ultrasonic seed detection, experiments have been conducted using a pure solution in which only target seeds and precursor monomers exist (16,17,18). In contrast, biological samples of clinical interest are abundant in biomolecules, which are not the target of the assay but have a potential impact on seed detection (21). These biomolecules can alter seed detection by, for example, nonspecifically binding to seeds and forming complexes that are incompetent to grow as fibrils (22,23,24,25).

In this study, we investigate the effects of ultrasonication on the $\alpha$-Syn seed detection from a highly crowded sample. To mimic a crowded environment, we use human serum albumin (HSA), a highly abundant protein in blood (26), as a crowding agent. Before the seed detection assay, the amyloid formation of $\alpha$-Syn monomers in a crowded milieu is investigated to clarify the general effects of the crowding agent. Subsequently, the ultrasonic seed detection assay of $\alpha$-Syn is performed, and the results obtained under ultrasonication are compared with those obtained using the conventional shaking method, revealing a substantial difference in seed-dependent amyloid formation kinetics. To reveal the biophysical mechanism behind the obtained results, we further discuss the mechanism of the seed-dependent amyloid formation by analyzing the interactions among the $\alpha$-Syn monomer, $\alpha$-Syn seeds, and HSA using quartz crystal microbalance (QCM) biosensor and developing the theoretical model describing the amyloid formation kinetics based on the master equation approach.

Materials and methods

Materials

Recombinant $\alpha$-Syn monomers expressed using Escherichia coli (BL21-DE3, NIPPON GENE, Tokyo, Japan) were purified using liquid-phase chromatography, as previously described (27). The obtained $\alpha$-Syn monomers were lyophilized and kept at −20°C before the experiments. The lyophilized $\alpha$-Syn monomers were first dissolved in deionized water and filtered using a membrane filter with a pore diameter of 220 nm. The protein concentration was determined using optical absorbance at 280 nm. The monomer solution was then mixed with chemicals. Their final concentrations are as follows: [$\alpha$-Syn] = 0.1 mg/mL, [NaPi(pH 7.0)] = 20 mM, [NaCl] = 1000 mM, [thioflavin-T] (ThT) = 5 $\mu$M, and [HSA] = 0–40 mg/mL.

Ultrasonic amyloid formation assay

We used an originally developed ultrasonic equipment (16,28) that irradiates sample solutions in a microwell plate with ultrasound and measures the fluorescence intensity of sample solutions (Fig. 1 A). The 200 $\mu$L sample solutions were added to a 96-well plate (675096, Greiner, Monroe, North Carolina) and sealed with a plastic film (547-KTS-HC, Watson, Radnor, Pennsylvania). The PZT (Lead zirconate titanate, Pb(Zr_x, Ti_1−x)O₃) transducers were placed on individual sample solutions. The frequency of the ultrasound was set to $\sim 30$ kHz, which is the optimum frequency for accelerating the amyloid formation (29). Ultrasonication was performed with a duty cycle comprising 0.5-s irradiation and 29.5-s pose periods. The ThT dye, which fluoresces upon binding to amyloid structures (30), was used to monitor the time course of amyloid formation. The ThT fluorescence intensity of each sample was measured every 5 min with excitation and emission wavelengths of 450 and 492 nm, respectively. The sample solution temperature was kept at 37°C. We defined the lag time for amyloid formation under ultrasonication as the time when the ThT fluorescence intensity exceeds 250.

Figure 1.

(A) Schematic illustration of the laboratory-build ultrasonic amyloid inducer. (B) Representative time-course curves of ThT fluorescence intensity of $\alpha$-Syn sample solutions with various concentrations of HSA. All curves are shown in Fig. S2$\left(n=5\right)$. (C and D) HSA concentration dependency of (C) lag times and (D) ThT maximum intensity of $\alpha$-Syn samples. The plots and error bars denote row values and standard deviation among five independent samples, respectively. (E) Circular dichroism (CD) spectra of the $\alpha$-Syn monomer without HSA before and after 20-h ultrasonication. (F) Transmission electron microscopy (TEM) image of ThT-positive $\alpha$-Syn aggregates formed by ultrasonication. The scale bar denotes 500 nm. (G) High-performance reverse-phase chromatograms for HSA before and after ultrasonication.

Shaking amyloid formation assay

We used a microplate reader (SH-9000, Corona Electric, Hitachi, Japan) for the assay under shaking agitation. The sample solutions (100 $\mu$L) were dispensed into the 96-well plate. The 96-well plate was shaken at 850 rpm with the cycle of 1 min shaking and 9 min incubation, which is typically adopted in conventional shaking assays (5). The ThT fluorescence intensity of each sample solution was measured every 10 min with excitation and emission wavelengths of 450 and 492 nm, respectively. The sample solution temperature was kept at 37°C. We defined the lag time for amyloid formation under shaking as the time when the ThT fluorescence exceeds 1000.

Seeding experiments

For seeding experiments, the amyloid fibrils of $\alpha$-Syn were preformed by ultrasonication at a monomer concentration of 0.1 mg/mL ([NaPi(pH 7.0)] = 20 mM, [NaCl] = 1000 mM, and [ThT] = 5 $\mu$M) under the condition without HSA. The obtained fibrils were broken into shorter fragments using an ultrasonic homogenizer (XL-2000, Misonix, Farmingdale, New York) by irradiating the seed solution with a 1-s ultrasonic pulse 10 times and diluted to intended concentrations with 20 mM sodium phosphate buffer. The seeds were added to the monomer solutions([$\alpha$-Syn] = 0.1 mg/mL, [NaPi(pH 7.0)] = 20 mM, [NaCl] = 1000 mM, and [ThT] = 5 $\mu$M) at various seed concentrations. It should be noted that amyloid fibrils can dissolve into monomers when the concentration of the seeds is lower than their solubility. After dilution, the seeds were promptly added to avoid the dissolution of fibrils. The prepared seeds were stored at room temperature to prevent cold denaturation of $\alpha$-Syn amyloid fibrils (27,31).

Circular dichroism spectrum measurements

Far-UV circular dichroism spectra (200–250 nm) were obtained by a spectropolarimeter (J-820, Jasco, Hachioji, Tokyo). The measurements were performed at 20°C using a quartz cell with a 1-mm path length.

Transmission electron microscopy observation

The 3 $\mu$L aliquot was placed onto a copper grid for 30 s and removed with filter paper. The sample was stained by 3 $\mu$L uranyl acetate for 5 s. The samples were visualized by a transmission electron microscope (JEM-1400, JEOL, Akishima, Japan) with an acceleration voltage of 80 kV. The observation was performed with a magnitude of 8000.

High-performance reversed-phase chromatography

A solution containing HSA with a concentration of 40 mg/mL after ultrasonication was collected and diluted with 20 mM NaPi buffer to 1 mg/mL. For the analysis, we used a high-performance liquid-phase chromatography system (GILSON, Middleton, Wisconsin) with a C4 300-Å column (5C4-AR-300, COSMOSIL, Houston, Texas). The chromatographic data were collected using Clarity (v.7).

QCM measurements

We used the wireless-electrodeless QCM biosensor for the interaction analysis between different molecule species, as described previously (32,33,34). The AT-cut quartz chip with a fundamental resonant frequency of 65 MHz and an in-plane dimension of $2.6\times 1.6{\text{mm}}^2$ was used as a resonator for the QCM measurement. We prepared two types of quartz resonators with and without a 2-nm chromium layer and a 15-nm gold layer on the surface of the resonator for analyzing interactions between the $\alpha$-Syn monomer and HSA and between the $\alpha$-Syn monomer and seeds, respectively.

For the analysis of the interaction between the $\alpha$-Syn monomer and HSA, the quartz resonators were initially rinsed by a piranha solution (98% ${\mathrm{H}}_2{\text{SO}}_4$:30% ${\mathrm{H}}_2{\mathrm{O}}_2$ = 7:3), followed by a rinse with deionized water. A self-assembled monolayer was formed by incubating the resonators in a 10-mM 10-carboxy-1-decanethiol solution prepared in anhydrous ethanol at 4°C overnight. After washing with ethanol and deionized water, a mixture of 100 mM N-hydroxysuccinimide and 100 mM 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride in deionized water was applied to activate the self-assembled monolayer termini. This activation step lasted for 2 h at room temperature, followed by rinsing with deionized water. To immobilize the $\alpha$-Syn monomer on the resonator surface, the resonator was soaked in 150 $\mu$M $\alpha$-Syn monomer solution with 20 mM NaPi buffer (pH 7.4) for 2 h at room temperature. After immobilization, solutions containing various concentrations of HSA were flowed using a micropump (APP-20KG, Takasago, Tokyo, Japan). The resonant frequency of the resonator was monitored using a laboratory-built setup while flowing the HSA solution with a flow rate of $\sim$0.5 mL/min at 37°C. The time course of the resonant frequency was recorded via a custom-developed software program.

For the analysis of the interaction between the $\alpha$-Syn monomer and seeds, after cleaning the resonators with piranha solutions, 3-aminopropyltriethoxysilane (Sigma-Aldrich, St. Louis, Missouri) was spotted onto the resonator to functionalize the surface of the resonator and incubated for 120 s (35), followed by rinsing with buffer solution. Then, 0.01 mg/mL $\alpha$-Syn seeds diluted with 20 mM NaPi buffer were placed on both sides of the resonators overnight at room temperature. After immobilization, solutions containing various concentrations of $\alpha$-Syn monomer were flowed using a micropump.

Data analysis of QCM data

The frequency response data were fitted with an exponential function of $\Delta f/f_0=A\left({\mathrm{e}}^{-\alpha t}-1\right)$ to extract the fitting constant $\alpha =k_{\text{on}}^{\mathrm{A}}\left[\text{HSA}\right]+k_{\text{off}}^{\mathrm{A}}$ (36), which are the rate constants for the association and dissociation reactions. The $\alpha$ values were then plotted against the HSA concentration to determine $k_{\text{on}}^{\mathrm{A}}$ and $k_{\text{off}}^{\mathrm{A}}$, which correspond to the slope and intercept of the linear regression line, respectively. The dissociation constant $\left(K_{\mathrm{D}}^{\mathrm{A}}\right)$ was calculated using $K_{\mathrm{D}}^{\mathrm{A}}=k_{\text{on}}^{\mathrm{A}}/k_{\text{off}}^{\mathrm{A}}$, assuming a 1:1 binding stoichiometry between HSA and $\alpha$-Syn monomers.

Time-evolution simulation of $\alpha$-Syn amyloid formation

To theoretically describe the time evolution of fibril formation at varying initial seed concentrations based on nucleation, fibril growth, and fragmentation reaction pathways, we apply the master equation approach as proposed by Knowles and colleagues (37,38),

$$\frac{\partial C_F\left(t,j\right)}{\partial t}=2k_{\text{on}}^{\mathrm{F}}{C}_F\left(t,j-1\right)M\left(t\right)-2k_{\text{off}}^{\mathrm{F}}{C}_F\left(t,j\right)+2k_{\text{off}}^{\mathrm{F}}{C}_F\left(t,j+1\right)+2k_{-}\sum _{k=j+2}^{\infty }{C}_F\left(t,k\right)+{\delta }_{j,n_c}{k}_n{C}_M{\left(t\right)}^{n_c}\text{,}$$ (1)

where $C_F\left(t,j\right)$ represents the molar concentration of fibrils of length $j$ at time $t$ and $C_M\left(t\right)$ denotes the molar concentration of the $\alpha$-Syn monomer at time $t$. $k_{\text{on}}^{\mathrm{F}}$ and $k_{\text{off}}^{\mathrm{F}}$ represent the rate constants for the association and dissociation of the monomers at the fibril termini, respectively. $k_n$ and $k_f$ denote the rate constants for the primary nucleation and fragmentation pathways, respectively. Here, the critical nucleus size for primary nucleation, $n_c$, is set to $n_c=2$. The rate constants determined by QCM assays are given as $k_{\text{on}}^{\mathrm{F}}=6.7\times {10}^1{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$ and $k_{\text{off}}^{\mathrm{F}}=6.0\times {10}^{-5}{\mathrm{s}}^{-1}$. The $k_n$ and $k_f$ values are taken from previous literature (16,18), where ultrasonication was used as an agitation method, $k_n=1.0\times {10}^{-7}{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$ and $k_{-}=1.0\times {10}^{-7}{\mathrm{s}}^{-1}$.

Results

The effects of HSA on amyloid formation of $\alpha$-Syn monomer

We performed the amyloid formation assay under ultrasonication using the laboratory-built ultrasonic amyloid inducer in conjugation with the ThT fluorescence assay (16,39). The sample solutions, including the $\alpha$-Syn monomer, HSA, and amyloid-specific ThT dye, were placed in a 96-well microplate and sonicated by piezoelectric transducers with a fundamental resonant frequency of $\sim 30$ kHz (Fig. 1 A). It should be noted that our ultrasonic assay causes a temperature increase of less than 0.1°C in the sample solution due to the short ultrasonication duty cycle. Therefore, the effects of temperature change caused by ultrasonication can be considered negligible. As a control experiment, we also performed the amyloid formation assay under shaking agitation.

The amyloid formation of $\alpha$-Syn monomers under ultrasonication was investigated upon the addition of HSA with various concentrations (Figs. 1, B–D, and S1). Here, it should be noted that a blood concentration of HSA is typically $\sim 40$ mg/mL (40). Thus, we tested the effects of HSA with a concentration of up to 40 mg/mL. Without the addition of HSA, the ThT fluorescence intensity of the $\alpha$-Syn solution begins to increase after a lag time of 2.5 h and hits a plateau at 7.5 h (Figs. 1 B and S2). The circular dichroism spectra (Fig. 1 E) show a transition of the secondary structure of $\alpha$-Syn monomers from a random coil to a $\beta$-sheet-rich structure before and after the experiment, as indicated by the negative peak near 220 nm. Transmission electron microscopy visualizes the short fibrillar morphology of $\alpha$-Syn aggregates formed under ultrasonication (Fig. 1 F), showing that our ultrasonic assay produces $\alpha$-Syn amyloid fibrils.

When the HSA concentration is below 20 mg/mL, the addition of higher concentrations of HSA results in an extended lag time (Fig. 1 C) and a lower ThT maximum intensity (Fig. 1 D). We confirmed that there is no significant quenching effect of the addition of HSA on the ThT fluorescence intensity (Fig. S3). Therefore, this result indicates that HSA inhibits the amyloid formation of the $\alpha$-Syn monomer in a dose-dependent manner within this concentration range. However, when HSA concentration is above 20 mg/mL, the inhibitory effect of HSA appears to be independent of its concentration. Ahmed and co-workers reported that the inhibitory effect of HSA arises from the binding between the $\alpha$-Syn monomer and HSA, leading to the formation of an off-pathway complex from amyloid formation (41). Assuming that the molar ratio of the $\alpha$-Syn monomer and HSA is approximately 1:40 when the HSA concentration is 20 mg/mL ([$\alpha$-Syn] = 6.9 $\mu$M and [HSA] = 301 $\mu$M), an excess addition of HSA does not show significant effects on the inhibition of $\alpha$-Syn amyloid formation above an HSA concentration of 20 mg/mL.

Ultrasonication has the potential to chemically modify biomolecules by generating radical species (42,43,44). Then, we examined whether the structure of HSA changes before and after ultrasonication using the reversed-phase liquid chromatography (Fig. 1 G). The elution peak of HSA shows no significant shift, indicating that intact HSA remains after ultrasonication.

Control experiments using conventional shaking agitation (Fig. S4) also reveal the inhibitory effects of HSA on the amyloid formation of the $\alpha$-Syn monomer. These findings are consistent with previous studies (41,45), which reported that HSA interacts with $\alpha$-Syn monomers to reduce the fraction of $\alpha$-Syn monomers available for amyloid formation. Our results demonstrate that the ultrasonication amyloid formation assay compromises neither the inherent nature of $\alpha$-Syn amyloid formation nor the ability of HSA as a crowding agent.

Ultrasonic $\alpha$-Syn seed detection assay from a crowded environment

We investigated the effects of HSA on $\alpha$-Syn seed-dependent amyloid formation under ultrasonication (Fig. S1 B). Preformed $\alpha$-Syn fibrils with a concentration of 0.1 mg/mL were used as seeds. Here, the concentration range of the seeds was determined by referring to a study by Okuzumi and co-workers (6), which succeeded in distinguishing patients with PD from healthy individuals by combining RT-QuIC and immunoprecipitation, whose detection sensitivity was approximately 1 ng/mL. Then, we prepared sample solutions with seed concentrations ranging from 0.01 to 100 ng/mL.

In the absence of HSA, the addition of seeds with concentrations of 10 and 100 ng/mL enhances $\alpha$-Syn amyloid formation compared to the sample without seeds (Figs. 2 A and S5). However, at seed concentrations below 1 ng/mL, the acceleration effect by seeds becomes unclear, as shown by the overlapped ThT curves with that of the sample without seeds (Fig. 2 A). The ThT maximum intensities remain unchanged across all seed concentrations.

Figure 2.

(A–D) Representative time-course curves of ThT fluorescence intensity (A) without HSA and with (B) 40, (C) 20, and (D) 5 mg/mL HSA under ultrasonication. All curves for (B)–(D) are shown in Figs. S6, S7, and S8, respectively. (E and F) Seed concentration dependency of (E) lag times and (F) ThT maximum intensity for $\alpha$-Syn amyloid formation under ultrasonication. (G) Representative time-course curves of ThT fluorescence intensity in the presence of seeds with various concentrations with 40 mg/mL HSA under shaking. The legend colors are common to (A)–(D) and (G). All curves are shown in Fig. S9. (H) HSA concentration dependency of ThT maximum intensity for $\alpha$-Syn amyloid formation under shaking. The legend colors are common to (E), (F), and (H). Error bars in (E), (F), and (H) denote the standard deviation among independent sample solutions $\left(n=5\right)$.

Intriguingly, the effects of seeds are more pronounced in the crowded environment under ultrasonication (Figs. 2, B–D, and S6–S8). As the seed concentration increases, the lag time for the amyloid formation shortens (Fig. 2 E), and the ThT maximum intensity increases (Fig. 2 F). This tendency is consistent at HSA concentrations of 5, 20, and 40 mg/mL (Fig. 2, B–D). However, this tendency is not observed under shaking agitation (Figs. 2, G and H, and S9). These results indicate that ultrasonication enhances $\alpha$-Syn seed detection in a crowded milieu by a mechanism different from conventional shaking agitation.

Seeds can be detected based on differences in the ThT time-course kinetics, specifically the ThT maximum intensity and the lag time, between samples with and without seeds. Here, we evaluate the seed detection sensitivity using both the lag time and ThT maximum intensity. Seed detection sensitivity is defined as the lowest seed concentration at which the distribution of lag times or ThT maximum intensities (mean $\pm$ standard deviation) does not overlap with that of samples without seeds, which is a negative control for the assay (Fig. S10). The detection sensitivities of the seeds with 40 mg/mL HSA based on the lag time and ThT maximum intensity are 1 and 0.01 ng/mL, respectively, showing that the ThT fluorescence intensity is a more sensitive index than the lag time in the ultrasonic assay. In contrast, the conventional shaking method fails to detect $\alpha$-Syn seeds upon the addition of HSA, highlighting the effectiveness of ultrasonication for detecting seeds from crowded environments.

Ultrasonication has the ability to enhance primary nucleation (29,46), which is spontaneous nucleation from monomers, besides its robust ability of fragmentation. In the context of seed detection, the ultrasonic enhancement of primary nucleation is not preferable because it can result in a false positive by producing seed-independent fibrils. In pure solutions without HSA (Fig. 2 A), ultrasonication induces spontaneous nucleation, masking the acceleration effects of seeds. However, in a crowded solution with HSA, especially in the case of 40 mg/mL HSA, the sample without seeds shows no significant increase in the ThT fluorescence intensity under ultrasonication. We further investigate the correlation between the ThT fluorescence intensity and the amount of resultant amyloid fibrils (Fig. S11). The results indicate a positive correlation between the maximum ThT fluorescence intensity and the amount of resultant fibrils, suggesting that lower seed concentrations under ultrasonication lead to reduced fibril yields. In other words, the formation of the seed-independent fibrils is suppressed under ultrasonication. Because this is not observed in the case of shaking, the suppression of the seed-independent amyloid formation in a crowded environment is a specific phenomenon for ultrasonication. Furthermore, this suggests that the detection sensitivity of seeds depends on the type of agitation. The reduction in seed-independent fibril formation lowers the possibility of false positives in a seed detection assay, thereby improving the performance of seed detection.

From the physicochemical perspective, our experimental result that the initial seed amounts determine the resultant fibril amounts raises an important question. Amyloid formation is a phase transition of supersaturated denatured protein monomers, analogous to the crystallization of small compounds such as salts (47,48,49,50). The amount of the resultant fibrils should be determined by the amount of monomers exceeding their thermodynamic solubility. The addition of seeds can ignite the phase transition of denatured monomers into amyloid fibrils by acting as a template but should not affect the resultant amyloid amount. To resolve this contradiction and understand the role of ultrasonication in seed detection from a crowded milieu, we further analyze and discuss the biophysical mechanisms behind ultrasonic seed detection in the following sections.

Analysis of interaction among $\alpha$-Syn monomer, seed, and HSA

To elucidate the mechanism underlying the correlation between the amount of resultant fibrils and the initial seed concentration under ultrasonication, we investigate the interactions among the $\alpha$-Syn monomer, seed, and HSA using the originally developed QCM biosensor (32,33). Briefly, QCM measurements are based on the relationship between the mass of a quartz resonator and its resonant frequency: an increase in mass results in a decrease in the resonant frequency (51). Thus, the adsorption of molecules onto the resonator surface reduces the resonant frequency by increasing its mass. By immobilizing a molecule on the resonator surface, injecting another molecule, and monitoring the frequency shift during the molecular interaction, we can determine the dissociation constant through an analysis of the frequency response (36).

First, $\alpha$-Syn monomers are immobilized on the QCM sensor, and various concentrations of HSA are injected to investigate the interaction between the $\alpha$-Syn monomer and HSA (Fig. 3 A). After injection of HSA, the resonant frequency of the resonator decreases in a concentration-dependent manner (Fig. 3 B). Here, we hypothesize that the binding between $\alpha$-Syn monomers and HSA occurs in a 1:1 stoichiometry and analyze the reaction curve by fitting the theoretical curve to extract the dissociation constant, $K_{\mathrm{D}}^{\mathrm{H}}$ (Fig. 3 C). The dissociation constant between the $\alpha$-Syn monomer and HSA is $K_{\mathrm{D}}^{\mathrm{H}}=417$ $\mu$M, indicating the weak interaction between them. This value is consistent with the previous study by Ahmed and co-workers (41), which reported the dissociation constant of $520\pm 12$ $\mu$M. The association and dissociation rate constants ($k_{\text{on}}^{\mathrm{H}}$ and $k_{\text{off}}^{\mathrm{H}}$) are determined to be $1.7\times {10}^1{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$ and $7.0\times {10}^{-3}{\mathrm{s}}^{-1}$, respectively.

Figure 3.

(A) Experimental scheme of QCM for analyzing the interaction between $\alpha$-Syn monomer and HSA. The $\alpha$-Syn monomer was immobilized on the QCM sensor and flowed through various concentrations of HSA solutions. (B) Change in the resonant frequency of a QCM chip by the injection of HSA and (C) calculation of the dissociation constant, $K_{\mathrm{D}}^{\mathrm{H}}$, between $\alpha$-Syn monomer and HSA. The error bars denote the standard deviation among independent measurements $\left(n=3\right)$. (D) Experimental scheme of QCM for analyzing the interaction between $\alpha$-Syn monomer and $\alpha$-Syn seeds. Seeds were immobilized on the QCM sensor and flowed through various concentrations of $\alpha$-Syn monomer solutions. (E) Change in the resonant frequency of a QCM chip by the injection of $\alpha$-Syn monomer and (F) calculation of the dissociation constant, $K_{\mathrm{D}}^{\mathrm{F}}$.

Next, we investigate the interaction between the $\alpha$-Syn monomer and seed (Fig. 3 D). The results show a $K_{\mathrm{D}}^{\mathrm{F}}$ of 895 nM for the interaction between the $\alpha$-Syn monomer and seed (Fig. 3 E), which is a thousandfold higher value than that between the $\alpha$-Syn monomer and HSA. The $k_{\text{on}}^{\mathrm{F}}$ and $k_{\text{off}}^{\mathrm{F}}$ values are obtained as $6.7\times {10}^1{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$ and $6.0\times {10}^{-5}{\mathrm{s}}^{-1}$, respectively (Fig. 3 F). We hypothesize that these binding kinetics mainly reflect the elongation rate of each fibril, specifically involving monomer binding to fibril termini, although the observed kinetics may also include contributions from nonspecific binding of monomers to the resonator surface or lateral binding along the fibril surface. On the basis of this assumption, we use these experimentally obtained rate constants in the following kinetic simulation of $\alpha$-Syn amyloid formation in a crowded milieu.

Time-evolution simulation of seed-dependent amyloid formation in a crowded milieu

We develop a theoretical model to describe seed-dependent amyloid formation, incorporating the effects of HSA as a molecular crowder and ultrasonication. This model builds upon previous studies on the kinetics of nucleated polymerization (37,38,52). Specifically, we consider a link between amyloid formation under ultrasonication and the complex formation between $\alpha$-Syn monomers and HSA molecules.

To explain amyloid formation behavior under ultrasonication, we consider three reaction pathways: 1) primary nucleation, in which soluble monomers associate to form the fibril nuclei (Fig. 4 Ai); 2) fibril elongation, in which monomers bind to the fibril termini, leading to fibril growth (Fig. 4 Aii); and 3) fibril fragmentation, in which a single fibril breaks into multiple shorter fragments (Fig. 4 Aiii). Although it has been reported that oligomer formation (53) and secondary nucleation (54) are involved in the $\alpha$-Syn amyloid formation process, fragmentation is particularly pronounced during ultrasonically induced amyloid formation (16,20). Therefore, we adopt a fragmentation-dependent model to describe amyloid formation under these conditions and calculate it based on the master equation approach, as described in the materials and methods section.

Figure 4.

Reaction schemes of (A) $\alpha$-Syn amyloid formation and (B) HSA complex formation. M, F, H, R, and IR stand for $\alpha$-Syn monomer, $\alpha$-Syn fibril, HSA monomer, reversible complex, and irreversible complex, respectively. (C and D) The time course of formed fibril concentrations with various initial seed concentrations by simulation that considers (C) reversible complex formation (scheme i in B) and (D) irreversible complex formation in addition to the reversible one (schemes i–iii in B), respectively. (E) Initial seed concentration dependence of the final concentrations of $\alpha$-Syn monomers involved in four different species in the result of the simulation shown in (D). Detailed parameters are shown in Table S1. The legend colors are common to (C) and (D).

Regarding the HSA complex formation (Fig. 4 Bi), we first consider a complex composed of a single $\alpha$-Syn monomer and an HSA molecule, which is a reversible complex that can dissociate dynamically,

$$\frac{d{C}_R\left(t\right)}{dt}=k_{\text{on}}^{\mathrm{H}}M\left(t\right)H\left(t\right)-k_{\text{off}}^{\mathrm{H}}{C}_R\left(t\right)\text{.}$$ (2)

Here, $C_R\left(t\right)$ represents the molar concentration of the reversible $\alpha$-Syn-HSA complex at time $t$ and $H\left(t\right)$ denotes the molar concentration of the free HSA molecule at time $t$. $k_{\text{on}}^{\mathrm{H}}$ and $k_{\text{off}}^{\mathrm{H}}$ are the rate constants for the association and dissociation of the $\alpha$-Syn monomer and HSA molecule, respectively. We use the experimentally obtained values for these rate constants: $k_{\text{on}}^{\mathrm{H}}=1.7\times {10}^1{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$ and $k_{\text{off}}^{\mathrm{H}}=7.0\times {10}^{-3}{\mathrm{s}}^{-1}$.

By linking Eqs. 1 and 2, we calculate the time evolution of $\alpha$-Syn amyloid formation in the presence of HSA molecules under different initial seed concentrations (Fig. 4 C). The simulation conditions are set to match those of the experiment shown in Fig. 2 C, where the HSA concentration is fixed at 40 mg/mL and the initial seed concentration varies from 0.01 to 100 ng/mL. Notably, the fibril concentration $F\left(t\right)$ on the vertical axis of Fig. 4 C represents the monomer-equivalent concentration of formed fibrils at time $t$, calculated as $F\left(t\right)={\sum }_{j=2}^{\infty }j{C}_F\left(j,t\right)$. Although the simulation shows the seed concentration-dependent acceleration of fibril formation, it fails to simulate the experimental observation that the final fibril yield depends on the initial seed concentration (Fig. 2 B). Simulation of the time evolution of free $\alpha$-Syn monomer and HSA concentrations (Fig. S12) reveals that these two species reach equilibrium within $\sim 300$ s after the beginning of the reaction. This rapid equilibrium suggests that the effects of the initial seed concentration on the resultant fibril yield cannot be explained solely by the formation of reversible complex formation between the $\alpha$-Syn monomer and HSA.

To account for this discrepancy, a second complex distinct from the first one is introduced (Fig. 4, Bii and Biii). Here, we hypothesize that the second complex is an irreversible complex formed through the higher-order aggregation of the reversible complexes. The formation process of the irreversible complex is expressed as

$$\frac{d{C}_{IR}\left(t\right)}{dt}=k_{\mathrm{c}}{C}_R\left(t\right)\left(\frac{1}{2}{C}_R\left(t\right)+C_{IR}\left(t\right)\right)\text{.}$$ (3)

Here, $C_{IR}\left(t\right)$ represents the molar concentration of the irreversible complex at time $t$ and $k_{\mathrm{c}}$ represents the rate constant for the formation of the irreversible complex, and this is set to $k_c=9.5$ ${\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$.

By linking Eqs. 1, 2, and 3, we performed a time-evolution simulation of $\alpha$-Syn fibril formation in the presence of 40 mg/mL HSA and various initial seed concentrations (Fig. 4 D). The results indicate that the second model, which includes the irreversible $\alpha$-Syn-HSA complex, successfully captures the dependence of the final fibril yield on the initial seed concentration. Fig. 4 E shows the $\alpha$-Syn monomer concentrations in different species at various initial seed concentrations. A higher initial seed concentration leads to an increased concentration of $\alpha$-Syn monomers incorporated into the fibrils and a reduced concentration of $\alpha$-Syn monomers in the irreversible complex.

Discussion

Theoretical calculations, considering two distinct complexes formed between the $\alpha$-Syn monomer and HSA, successfully capture the experimentally observed trend that the final fibril yields depend on the initial seed concentrations. Fig. 5 illustrates the thermodynamic energy landscape of amyloid formation under ultrasonication in the presence of HSA as a crowding agent.

Figure 5.

Energy landscape of the aggregation reaction with different seed concentrations under ultrasonication. The colors of the curves denote the corresponding seed concentration qualitatively, as shown in the inset.

The $\alpha$-Syn monomers involved in the reversible complex can dynamically dissociate due to the high dissociation constant ($K_D^H=417\mu$M). In the absence of the irreversible complex, the final fibril yield is governed by the thermodynamic equilibrium among three molecular species: $\alpha$-Syn monomers, fibrils, and the reversible $\alpha$-Syn-HSA complex. This equilibrium is primarily dictated by two dissociation constants, $K_D^H$ and $K_D^F$, which determine the equilibrium among the aforementioned three species. Consequently, in this reaction system, the initial seed concentration does not influence the final fibril yield.

The presence of the irreversible complex introduces kinetic effects into the reaction scheme. $\alpha$-Syn monomers incorporated into the irreversible complex cannot dissociate back into free monomers, leading to a kinetic competition between irreversible complex formation and amyloid fibril formation. At higher initial seed concentrations, the increased number of fibril termini provides more binding sites for monomers, facilitating their rapid incorporation into fibrils. As a result, higher initial seed concentrations lead to a greater proportion of monomers being incorporated into fibrils, thereby increasing the final fibril yield (Fig. 5).

One possible reason for the peculiar formation of the irreversible complex under ultrasonication is molecular condensation induced by ultrasonic cavitation. When a liquid is subjected to ultrasonication, an oscillatory pressure field is generated. If the amplitude of this pressure field exceeds a certain threshold, the negative pressure induces the formation of bubbles, known as ultrasonic cavitation. These bubbles periodically expand and shrink in synchronization with the ultrasonic pressure field. The air-water interface of the cavitation bubbles can attract the protein molecules during its expansion phase because proteins tend to accumulate at the air-water interface via hydrophobic interaction (55). When bubbles shrink, the protein molecules adsorbed onto the bubble surface are concentrated at the collapse point of the bubble (Fig. 5). This results in the local condensation of the protein molecules, leading to the higher-order irreversible aggregates between the $\alpha$-Syn monomer and HSA. Because this phenomenon originates from ultrasonic cavitation, the initial seed concentration dependency of the final fibril yield is observed only under ultrasonication, not under shaking conditions.

In the context of seed detection, the ultrasonic formation of an irreversible complex between the target protein and the crowding agent offers distinct advantages. The kinetic competition between fibril formation and the formation of this irreversible complex leads to the observed dependency of the final fibril yield on the initial seed concentration. The final fibril yields are quantified as the maximum ThT fluorescence intensity, which has been shown to be a more sensitive indicator than the traditionally used lag time for amyloid formation. Currently, clinical seed detection involves measuring differences in lag times between samples from patients and healthy individuals. However, establishing experimental conditions that generate significant kinetic differences at ultralow seed concentrations remains a challenge. Furthermore, in biological samples, the termini of seeds are likely to be veiled by crowding proteins, potentially inactivating their ability to promote fibril formation. In such cases, ultrasonication can help expose active termini by fragmenting the seeds. This study demonstrates the potential for equilibrium-based seed detection using ultrasonication in a crowded environment, which could contribute to more reproducible and reliable seed detection methods.

In the context of seed quantification, our method still needs to be improved. To quantify the seed concentration by the maximum ThT fluorescence intensity, we need to obtain some calibration curve (i.e., testing line) to relate the resultant ThT kinetics to the seed concentrations included in the samples. However, the shape of a calibration curve could be influenced not only by amyloid seeds but also by varying matrix components, such as HSA, which can range widely between individuals and disease states. Furthermore, other matrix components beyond HSA may also affect the interaction and seeding kinetics of $\alpha$-Syn, and this complexity further underlines the need for orthogonal, quantitative methods. Therefore, our ultrasonication-based assay is best suited for rapid screening purposes, leveraging the robust amplification capability of ultrasonication to detect the presence of amyloid seeds. However, for the precise quantification of seed concentrations, which will be crucial for monitoring disease progression or evaluating therapeutic efficacy, alternative methods are necessary. In this regard, digital-based assays (56,57) hold strong potential for accurate quantification in complex matrices. Even if such digital assays become available and widely used, our rapid ultrasonication-based screening will remain valuable to identify potential patients early, thereby expanding diagnostic reach and enabling timely follow-up testing.

Conclusion

In this study, we developed a sensitive $\alpha$-Syn seed detection assay utilizing ultrasonication in a highly crowded environment. Our findings demonstrate that the ThT maximum intensity, which positively correlates with fibril yield, serves as a reliable indicator for $\alpha$-Syn seed detection. This is attributed to the dependence of the final fibril yield on the initial seed concentration under ultrasonication. Additionally, time-evolution simulations of $\alpha$-Syn fibril formation, incorporating two distinct $\alpha$-Syn-monomer-HSA complexes, successfully reproduced this phenomenon. The simulations revealed that the presence of an irreversible complex induced kinetic competition between amyloid formation and complex formation, leading to the initial seed concentration dependency of the final fibril yield. Consequently, our ultrasonic assay achieved a detection limit of less than 10 pg/mL $\alpha$-Syn seeds in the presence of 40 mg/mL HSA, demonstrating its potential applicability for detecting $\alpha$-Syn seeds in biological samples. These findings underscore the adaptability of ultrasonication for efficient amplification and sensitive detection of $\alpha$-Syn seeds, contributing to the advancement of clinical diagnostics for PD and other synucleinopathies.

Acknowledgments

This study was supported by the JSPS (JP22K14013, JP24KK0104, and JP25K00016 to K.N.), JKA and its promotion funds (2024M-583 to K.N.), and JST SPRING (JPMJSP2138 to T.O.). The authors are grateful to the Daicel Corporation and the Research Center for Ultra-High Voltage Electron Microscopy of Osaka University.

Author contributions

T.O. and K.N. designed the research. T.O., K.N., and K.Y. carried out the experiments, simulations, and data analysis. T.O., K.N., Y.G., and H.O. discussed the results. T.O., K.N., and H.O. wrote the manuscript.

Declaration of interests

The authors declare no competing interests.

Editor: Samrat Mukhopadhyay