In the human body, a significant presence of intrinsically disordered proteins (IDPs) plays diverse and crucial roles (Fuxreiter and Tompa, 2012; Forman-Kay and Mittag, 2013; Bah and Forman-Kay, 2016). These proteins lack a well-defined 3D structure under native conditions, which imparts functional advantages, but also renders them susceptible to irreversible aggregation, especially when affected by mutations. Such aggregates can be pathogenic and are associated with various diseases, including neurodegenerative diseases, cancer, diabetes, and cardiovascular diseases (Uversky et al., 2008).

Notably, Alzheimer’s disease is characterized by the aggregation of the amyloid-β peptide (Aβ), while Parkinson’s disease (PD) is linked to α-synuclein (αS) aggregation. A growing body of evidence has established a connection between IDPs and the phenomenon known as liquid-liquid phase separation (LLPS). During LLPS, high and low concentrations of biomolecules coexist without the presence of membranes and exhibit properties similar to phase-separated liquid droplets of two immiscible liquids (Figure 1; Xing et al., 2021; Ray et al., 2020; Shu et al., 2021; Rodríguez et al., 2023; Gui et al., 2023). This intriguing phenomenon has garnered significant attention as it underlies the formation of membrane-less subcellular compartments (Hyman et al., 2014; Banani et al., 2017; Shin and Brangwynne, 2017), which, when dysregulated, can lead to incurable pathogenic diseases.

Figure 1

Download asset Open asset

Recent findings have highlighted the capability of αS to undergo LLPS under physiological conditions, specifically when the protein concentration surpasses a critical threshold (Ray et al., 2020). Moreover, it was observed that the aggregation propensity of αS is significantly influenced by various factors, including the presence of molecular crowders, the ionic strength of the protein environment, and pH (Sawner et al., 2021). Nonetheless, characterizing the interactions and dynamics of these small aggregates poses experimental challenges, leading to limited available reports on the subject (Apetri et al., 2006; Hong et al., 2011; Chen et al., 2015; Cremades et al., 2017).

This investigation aims to establish the molecular basis of self-aggregation of αS and underlying process of LLPS under diverse environmental perturbations. In particular, to understand the influence of environmental factors on the inter-protein interactions within a phase-separated droplet, we target to computationally simulate the aggregation process of αS under different conditions, emphasizing the roles of crowders and salt. While recent progress in computational force fields and hardware has enabled the simulation of individual IDPs especially αS, using all-atom molecular dynamics (AAMD) (Ahmed et al., 2021; Bari and Prakashchand, 2021; Robustelli et al., 2018; Best et al., 2014; Huang et al., 2017; Menon and Mondal, 2023; Menon and Mondal, 2022), these simulations can be extremely time-consuming and resource-intensive, making multi-chain AAMD simulations, even with cutting-edge software and hardware, impractical. Therefore, to simulate the the aggregation process, we resort to coarse-grained molecular dynamics (CGMD) simulations. Multiple CG force fields have been developed with the sole purpose of fast and accurate simulations of IDPs and LLPS (Dignon et al., 2018; Regy et al., 2021; Joseph et al., 2021; Tesei and Lindorff-Larsen, 2022). However these are implicit water, residue-level CG models. Therefore, here we leverage a tailored Martini 3 CG force field (CGFF) (Souza et al., 2021) for αS and use it to dissect the inter-protein interactions governing stable aggregate formation and LLPS. By leveraging the CGFF framework and building upon the groundwork laid by prior studies (Benayad et al., 2021; Thomasen et al., 2022; Mukherjee et al., 2023), we have optimized water-protein interactions for αS. Our multi-chain microsecond-long CGMD simulations have resulted in comprehensive ensembles of significant protein aggregates spanning various scenarios.

As one of the key observations, our simulation unequivocally reveals LLPS-like attributes in the aggregates and shows how these get modulated in the presence of crowders and salt. The investigation unearths the intricate interplay of mechanical and thermodynamic forces in αS aggregation, achieved through meticulous data analyses. We elucidate the pivotal intra- and inter-protein interactions governing LLPS-like protein droplet formation, unveiling the protein’s primary sequence’s role in aggregation. As would be shown in this article, a graph-based depiction of the droplet’s architecture represents the proteins within droplets as constituting dense networks akin to small-world networks.

In this study, we utilized the recently developed Martini 3 (Souza et al., 2021) CG model to simulate collective interaction of a large number of αS chains in explicit presence of aqueous media at various concentrations commensurate with in vitro conditions including the presence of crowders and salt. As Martini 3 was not originally developed for IDPs, we carefully optimized the protein-water interactions against atomistic simulation of monomer and dimer of αS, as detailed in the Methods section, to ensure compatibility with αS (see Methods).

Initially, we examined the impact of concentration on the protein’s aggregation by simulating copies of chains, maintaining a polydispersity of protein conformations of αS. In particular, three different conformations of αS (referred to here as ${\displaystyle ms1}$, ${\displaystyle ms2}$, and ${\displaystyle ms3}$) with R_gs (radius of gyration) ranging between collapsed and extended states (1.84–5.72 nm) at different concentrations, with a composition, as estimated in a recent investigation (Menon and Mondal, 2023), were employed. First the chains were simulated for extensive period in a set of three protein concentrations, close to previous experiments.

Simulations capture enhanced aggregation beyond a threshold concentrations of αS

We performed simulations of αS at various concentrations, namely 300 μM, 400 μM, 500 μM, and 750 μM. We begin by analyzing the aggregation behavior of αS. As shown in Figure 2, we observe that most chains do not aggregate at 300 and 400 μM as characterized by the prevalence of high number of free monomers. The respective snapshots of the simulation indicate the presence of greater extent of single chains. Also, the chains that are not free form very small oligomers of the order of dimer to tetramer (Figure 2).

Figure 2

Download asset Open asset

However, upon increasing the concentration to 500 μM, which has also been the critical concentration reported for αS to undergo LLPS (Ray et al., 2020), we observe a sharp drop in the average number of free monomers in the system (Figure 2). The corresponding representative snapshot of the system also depicts a few higher-order aggregates, such as pentamers and hexamers, as well as most chains forming small oligomers. This can be understood from the value of the average number of chains present in the largest clusters, as reported in Figure 2.

The system, being at critical concentration, formation of large aggregates would require longer timescales than the simulation length. Therefore, in order to promote the formation of large aggregates (heptamers or more) for finer characterization, we performed a simulation at a higher concentration of 750 μM αS. As shown in Figure 2d, we observe further decrease in the total number of free monomeric chains in the solution. There is simultaneous appearance of a very few droplet-like aggregates (hexamer or more) as can be seen from Figure 2 and the adjacent snapshot of the system (Figure 2). However, we note that ∼60% of the protein chains are free and do not participate in aggregation and we think that as such in water, αS does not possess a strong and spontaneous self-aggregation tendency. In the following sections we characterize the aggregation tendency of αS in the presence of certain environmental modulator that can shed more light on this hypothesis.

Molecular crowders and salt accelerate αS aggregation

The cellular environment, accommodating numerous biological macromolecules, poses a highly crowded space for proteins to fold and function (Ellis and Minton, 2006; Deeds et al., 2007; Li et al., 2008; Zhou, 2013). In in vitro studies, inert polymers such as Dextran, Ficoll, and polyethylene glycol (PEG) are commonly employed as macromolecular crowding agents. In the context of αS aggregation, previous experimental studies have revealed an increased rate of in vitro fibrillation in the presence of different crowding agents (Uversky et al., 2002; Munishkina et al., 2008; Horvath et al., 2021). Notably, a recent experimental study demonstrated the occurrence of phase separation (LLPS) of αS in the presence of PEG molecular crowder (Ray et al., 2020). Moreover, considering that in vivo environments also contain various moieties like salts and highly charged ions, a recent in vitro study has shown that the ionic strength of the solvent directly influences the aggregation rates of αS (Sawner et al., 2021), with higher ionic strength enhancing αS aggregation.

Given these observations, it becomes crucial to characterize the factors responsible for the enhanced aggregation of αS in the presence of crowders and salt. To address this, we perform two independent sets of simulations: one with αS present at 750 μM in the presence of 10% (vol/vol) fullerene-based crowders (see SI Methods) and the other with the same concentration of αS but in the presence of 50 mM of NaCl. In this section we characterize the effects of addition of crowders or salt on the aggregation of αS.

As expected, the addition of crowders leads to an enhancement of αS aggregation due to their excluded volume effects, as depicted in Figure 3a. Notably, the number of monomers drastically decreases upon the inclusion of crowders. This observation is further supported by the snapshots of the system, which also confirm the reduction in monomer count. Similarly, we observe that the presence of salt also promotes αS aggregation, as illustrated in Figure 3a, where the number of monomers is lower when compared to the case with no salt.

Figure 3

Download asset Open asset

Following this, we conducted an analysis of the number of chains present in the largest clusters that formed. Figure 3b clearly illustrates that the addition of crowder or salt leads to a notable increase in the average number of proteins forming a cluster. This crucial observation points to the fact that the inclusion of accelerators, such as crowder or salt, not only promotes aggregation but also plays a role in stabilizing the formed oligomers. Importantly, we observed that the effect of crowder on aggregation is slightly more pronounced compared to that of the salt. In the subsequent section, we delve into the reasons behind the enhanced aggregation induced by these accelerators, aiming to decipher the underlying mechanisms responsible for their influence on αS aggregation dynamics. As the aggregation is significant enough for performing quantitative analysis only when the concentration of αS is 750 μM, we perform all analysis on scenarios at 750 μM of αS.

Crowders and salt differentially modulate surface tension for promoting LLPS-like αS droplets

The preceding sections underscore our simulation-based observation that, influenced by crowders and salt, αS aggregates into higher-order oligomers (hexamers and beyond) at a significantly accelerated propensity compared to the scenario without these influences. Here, we delve into the investigation of the energetic aspects underlying this aggregation phenomenon. An important contributor to the energetics is the surface tension, arising from the creation of interfaces between the dense and dilute phases of the protein upon droplet formation. This presence of interfaces is accompanied by surface tension and surface energy. The surface energy of a system is directly proportional to its surface area; systems with higher surface energy tend to minimize their surface area. Consequently, systems comprising multiple smaller droplets exhibit a larger surface area, and hence a higher surface energy. Conversely, systems characterized by fewer, larger droplets possess a comparatively reduced surface area and correspondingly lower surface energy. This insight leads us to conjecture that surface tension could play a pivotal role in driving LLPS and the formation of larger αS droplets. To explore this hypothesis, we calculate the surface tension of the resultant droplets, as per Equations 1 and 2 and as described in SI Methods and Benayad et al., 2021.

(1) ${\gamma }_{20}=\frac{5k_{B}T}{16\pi ⟨(\delta a+\delta b{)}^2⟩}$

(2) ${\gamma }_{22}=\frac{15k_{B}T}{16\pi ⟨(\delta a-\delta b{)}^2⟩}$

where ${\displaystyle \delta a=a-R}$ and ${\displaystyle \delta b=b-R}$ is the perturbation of the droplet shape from a perfect sphere with a radius ${\displaystyle R}$ along any two pairs of principle axes of general ellipsoid estimating the shape of the droplet. The surface tension (${\displaystyle \gamma }$) is thus estimated using ${\displaystyle \gamma \approx {\gamma }_{20}\approx {\gamma }_{22}}$. Please see SI Methods and Benayad et al., 2021, for more details.

Figure 4a provides a comparison of the surface tension (${\displaystyle \gamma }$), for three different scenarios involving αS: (i) αS in solution, (ii) αS in the presence of crowders, and (iii) αS in the presence of salt. Notably, in each case, the surface tension is considerably lower (0.0035–0.0075 mN/m) than the surface tension for FUS droplets in water (∼0.05 mN/m) (Benayad et al., 2021). As stated earlier, the magnitude of surface tension is an estimate of the aggregation tendency of any liquid-liquid mixture. Since we find that γ_αS is much lower than ${\displaystyle {\gamma }_{FUS}}$, we assert that the propensity with which αS aggregates should be much lower than that of FUS.

Figure 4 with 1 supplement see all

Download asset Open asset

Next, we conduct a comparison of the three different scenarios to understand the effects of crowders and salt on the aggregation of αS. From Figure 4a, it is evident that the surface tensions are very similar for cases (i) and (ii), while it has increased for case (iii). This implies that the addition of crowders does not significantly impact the surface tension of the aggregates, although it renders the protein more prone to aggregation. On the other hand, the addition of salt causes an increase in surface tension. Given the relationship between surface area and volume, where a higher surface-to-volume ratio signifies numerous smaller droplets, the surface energy is concurrently elevated. In the presence of salt, a tendency is observed for these smaller aggregates to coalesce, giving rise to larger aggregates, albeit in reduced numbers. This behavior is an endeavor to curtail the surface-to-volume ratio and thus mitigate the associated surface energy. Therefore, the larger the surface tension, the higher is tendency of the protein to form aggregates, as seen from the surface tension values of αS and FUS, as mentioned earlier.

To minimize the surface energy, fusion of aggregates, either via merging of two or more droplets into one is seen for liquid-like phase-separated droplets in experiments (Ray et al., 2020). Although droplet fusion was not observed in our simulations due to the limited system size, it was shown that if a protein undergoes LLPS, a significant difference in protein concentration occurs between the droplet and the dilute phase (Nguyen et al., 2022). To verify whether the aggregates observed in our simulations exhibit characteristics of LLPS, we calculated the protein concentrations in the dilute and concentrated phases. For the droplet phase, the concentration of the protein was calculated using Equation 3.

(3) $c_{phase}=\frac{N_{phase}}{N_A\times V_{phase}}$

where ${\displaystyle N_{phase}}$ is the number of protein chains in the phase (here dilute or concentrated), ${\displaystyle N_A}$ is Avogadro’s number, and ${\displaystyle V_{phase}}$ is the volume occupied by the phase. For the dilute phase, we estimated the volume of the concentrated/dense phase (${\displaystyle V_{dense}}$) using Equation 4 (Nguyen et al., 2022).

(4) $V_{dense}^i=4\pi \sqrt{3}{\lambda }_1{\lambda }_2{\lambda }_3$

where ${\displaystyle V_{dense}^i}$ is the volume of the ${\displaystyle i}$th droplet, ${\displaystyle {\lambda }_1}$, ${\displaystyle {\lambda }_2}$, and ${\displaystyle {\lambda }_3}$ are the eigenvalues of the gyration tensor for the aggregate. The volume of the dilute phase is the remainder volume of the system given by Equation 5.

(5) ${\displaystyle V_{dilute}=V-\sum _i{V}_{dense}^i}$

where V is the total volume of the system.

As shown in Figure 4b and Figure 4—figure supplement 1, there is an almost two orders of magnitude difference between the concentration of αS in the dilute and droplet phases for all scenarios. Such a pronounced difference is a hallmark of LLPS, leading us to assert that the aggregates formed in our simulations possess LLPS-like properties. Consequently, we use the term ‘droplet’ interchangeably with ‘aggregates’ for the remainder of our investigation.

(6) $\mathrm{\Delta }{G}_{transfer}=RT\ln (\frac{c_{dilute}}{c_{dense}})$

Finally, utilizing the calculated concentrations, we proceed to estimate the excess free energy of monomer transfer (${\displaystyle \mathrm{\Delta }{G}_{transfer}}$), from Equation 6, between the dilute and droplet phases, where ${\displaystyle c_{dilute}}$ is the concentration of αS in the dilute phase, ${\displaystyle c_{dense}}$ is the concentration of αS in the dense/droplet phase, R is the universal gas constant, and T is the temperature of the system (=310.15 K). As illustrated in Figure 4c, both crowder and salt scenarios demonstrate lower ${\displaystyle \mathrm{\Delta }{G}_{transfer}}$ values compared to the case without their presence. However, the thermodynamic origins behind this pronounced aggregation differ for crowders and salt. Crowders enhance aggregation primarily through excluded volume interactions, which are of an entropic nature. On the other hand, salt enhances aggregation by increasing the droplet’s surface tension, thus contributing to the enthalpy of the system. As a result, apart from the already known fact that macromolecular crowding decreases ${\displaystyle \mathrm{\Delta }{G}_{transfer}}$ via entropic means, we also infer that salt decreases ${\displaystyle \mathrm{\Delta }{G}_{transfer}}$ via enthalpic means by increasing the surface tension of the formed droplets.

Aggregation results in chain expansion and chain reorientation in αS

An indicative trait of molecules undergoing LLPS is the adoption of extended conformations upon integration into a droplet structure. Given that the aggregates observed in our simulations exhibit a concentration disparity reminiscent of LLPS between the dilute and dense phases, we endeavored to validate the presence of a comparable chain extension phenomenon within our simulations (Nguyen et al., 2022). To address this, we quantified the radius of gyration (R_g) for individual chains and classified them based on whether they were situated in the dilute or dense phase. The distribution of R_g values for each category is illustrated in Figure 5a and Figure 5—figure supplement 1. Remarkably, the distribution associated with the dense phase distinctly indicates that the protein assumes an extended conformation within this context. As elucidated earlier, this marked propensity for extended conformations aligns with a characteristic hallmark of LLPS as previously seen in experiments (Ubbiali et al., 2022).

Figure 5 with 1 supplement see all

Download asset Open asset

Having observed the conformational alterations of αS during LLPS, our subsequent aim was to quantify the extent of these conformational changes in relation to their initial states (referred to as ‘ms1’, ‘ms2’, or ‘ms3’ in decreasing order of ${\displaystyle R_g}$; Menon and Mondal, 2023). To achieve this, we computed the root mean square deviation (RMSD) of each protein relative to its starting conformation. The resulting distributions were visually depicted using violin plots, featuring bold edges in Figure 5b and c. The protein ensemble was segregated into two categories: (i) those from the dilute phase (Figure 5b) and (ii) those from the dense phase (Figure 5c).

Surprisingly, regardless of their initial configurations, the observed RMSD values were notably high. To facilitate a comparative analysis, we also included distributions of RMSDs for single chains simulated in the presence of 50 mM of salt, depicted using violin plots with broken edges. Intriguingly, the conformational state labeled as ms1, exhibited the least RMSD, a characteristic attributed to its notably extended conformation. This phenomenon aligns with the preference of droplets for extended conformations, implying that ms1 required the least conformational perturbation and thus exhibited a lower RMSD.

For both ms2 and ms3, a conspicuous increase in RMSD values was observed across all proteins monomers, irrespective of their respective phases. This phenomenon can potentially be attributed to the pronounced conformational shift experienced by the protein during aggregation. Building on these observations, we put forward a hypothesis: LLPS engenders significant modifications in the native protein conformations, ultimately favoring the adoption of extended states.

As discussed in the previous paragraph that the αS monomers inside the droplets must undergo conformational expansion and we hypothesized that they adopt orientation so as to minimize the inter-chain electrostatic repulsions. To this end, we try to decipher the orientations of the chains via defining their axes of orientations and subsequently calculating the angles between the major axis of two monomers. We calculate the major axis of gyration, given by the eigenvector corresponding to the largest eigenvalue of the gyration tensor, for each monomer inside a droplet. We next find the nearest neighbor (minimum distance of approach <8 Å) for each monomer, carefully taking care of over-counting.

The angle between two monomers is defined as the angle between the major axes of gyration between chain ${\displaystyle i}$ and its nearest neighbor ${\displaystyle j}$. We plot the distributions of the angles for all scenarios and all droplets in Figure 5d. We observe that irrespective of the conditions, the distribution peaks at right angles. The representative snapshots (Figure 5e–h) showcase their mode of orientation. Interestingly the distribution is the same for all the three scenarios, again stressing upon the fact the αS droplets share similar features in terms of interactions and orientations irrespective of their environments.

Characterization of molecular interactions in aggregation-prone conditions

As established in preceding sections, both crowders and salt have been observed to augment the aggregation of αS while concurrently stabilizing the resultant aggregates. This phenomenon leads to the protein adopting extended conformations within a notably heterogeneous ensemble. Shifting our attention, we now delve into a residue-level investigation to unravel the specific interactions responsible for stabilizing these aggregates and, consequently, facilitating the aggregation process.

To compute the differential contact maps, our approach involved initial calculations of average intra-protein residue-wise contact maps, termed as intra-protein contact probability maps, for monomers present in both the dilute and dense phases (refer to Figure 6—figure supplement 1). Subsequently, we derived the difference by subtracting the contact probabilities of monomers within the dilute phase from those within the dense phase. As evident from Figure 6a, a discernible reduction in intra-chain Nter-Cter interactions is observed for monomers within the droplet phase, depicted by the presence of blue regions along the off-diagonals. Such a reduction in such interactions has also been observed via experiments (Ubbiali et al., 2022) and it is similarly noticeable in the two other cases, as evident in Figure 6b and c.

Figure 6 with 3 supplements see all

Download asset Open asset

Furthermore, a significant decline in intra-protein interactions, especially the NAC-NAC interactions, is predominantly observed at shorter ranges, indicated by deep-blue regions concentrated near the diagonals. Notably, these diminished intra-chain interactions (Figure 6 and Figure 6—figure supplement 1) potentially facilitate the formation of inter-chain interactions (Figure 6—figure supplement 2). Thus, we observed that increased inter-chain NAC-NAC regions (Figure 6—figure supplement 3) facilitate the formation of αS droplets which also have been previously seen from FRET experiments on αS LLPS droplets (Ray et al., 2020). Building on these observations, we posit that these interactions play a pivotal role in stabilizing the aggregates that have formed.

Moreover, from the difference heatmaps in Appendix 1—figure 5, it can be observed that the residues 95–110 (VKKDQLGKNEEGAPQE) have reduced contact probabilities upon introduction of crowders/salt, whereas the rest of the contacts have slightly increased. These residues are highly charged and we think that upon introduction of crowders/salt, the proteins inside the droplet needed to be spatially oriented to facilitate the formation of largest aggregates. This re-orientation occurs to minimize the electrostatic repulsions among these residues belonging to different chains. These analyses provide hints that these residues are present in the protein so as to avoid the formation of aggregation-prone conformations, which is why their interactions had to be minimized to form more stable and larger aggregates.

Phase-separated αS monomer forms small-world networks

The investigations so far suggest that irrespective of the factors that cause the aggregation of αS, the interactions that drive the formation of droplet remain essentially the same. However, the conformations of the monomers vary depending on their environment. In the presence of crowders they adapt to form much more compact aggregates. Therefore, here we characterize whether the environment influences the connectivity among different chains of the protein inside a droplet.

Figure 7a, b, and c shows molecular representations of the largest cluster formed by αS at 750 μM in water, αS at 750 μM in the presence of 10% (vol/vol) crowders, and αS at 750 μM in the presence of 50 mM NaCl, respectively. From the molecular representations for aggregates, it can be seen that irrespective of the system, they form a dense network whose characterization is not possible directly. Therefore, we represent each aggregate as a graph with multiple nodes (vertices) and connections (edges), as can be seen from Figure 7d, e, and f. Each node (in blue) represents a monomer in the droplet. Two nodes have an edge (line connecting two nodes) if the minimum distance of approach of the monomers corresponding to the pair of nodes is at least 8. We can see from the graph that not all chains are in contact with each other. They rather form a relay where a few monomers connect (interact) with most of the other protein chains. The rest of the chains have indirect connections via those. Since inter-chain connections/interactions have been denoted by edges and the chains themselves as nodes, such form of inter-chain interactions inside a droplet lead to only a few nodes having a lot of edges, e.g., node 1 in Figure 7f. The rest of them have only a few (3–5) edges. This is a signature of small-world networks (Watts and Strogatz, 1998; Barrat and Weigt, 2000; Humphries and Gurney, 2008; Farag et al., 2022) and we assert that αS inside the droplet(s) form small-world-like networks.

Figure 7

Download asset Open asset

A network can be classified as a small-world network by calculating the clustering coefficient and the average shortest path length for the network and comparing those to an equivalent Erdos-Renyi network (Rényi, 1959). The clustering coefficient (C) is a measure of the ‘connectedness’ of a graph, indicating the extent to which nodes tend to cluster together. It quantifies the likelihood that two nodes with a common neighbor are also connected. On the other hand, the average shortest path length (L) is a metric that calculates the average number of steps required to traverse from one node to another within a network. It provides a measure of the efficiency of information or influence propagation across the graph. To estimate the small-worldness of a graph, we calculate a parameter (${\displaystyle S}$) defined by Equation 7.

(7) $S=\frac{\frac{C}{C_r}}{\frac{L}{L_r}}$

where ${\displaystyle C}$ and ${\displaystyle L}$ are the clustering coefficient and average shortest path length for the graph generated for a droplet respectively, while ${\displaystyle C_r}$ and ${\displaystyle L_r}$ clustering coefficient and average shortest path length for an equivalent Erdos-Renyi network, respectively. Small-world networks exhibit the characteristic property of having ${\displaystyle C>>C_r}$, while ${\displaystyle L\approx L_r}$. In light of this, for every scenario (solely αS, αS in the presence of crowder, and αS in the presence of salt), we generate an ensemble of graphs that correspond to the droplets formed during the simulation.

For each graph, we calculate the small-worldness coefficient (${\displaystyle S}$) (Humphries and Gurney, 2008) and illustrate the distribution in Figure 7g. We observe a narrow distribution of ${\displaystyle S}$ with a mean of 3.4 for all cases. In a previous report of RNA-LLPS, a value of ${\displaystyle S\approx 4}$ was used to classify the droplets small-world networks (Nguyen et al., 2022). Therefore, ${\displaystyle S=3.4}$ would suggest that the droplets formed during the simulations are small-world like. Moreover, we observe that the distribution of ${\displaystyle S}$ is invariant with respect to the environment of the droplet.

Therefore, we establish that the modes of interactions, orientations, and even connectivities among αS monomers inside a droplet remain same even when their environments are extremely different. We think that this occurs since the residue-level interactions among different monomers inside the droplet are similar irrespective of the environment, as shown in a previous section. This puts forth a very interesting way of viewing αS LLPS. We think that if these residue-level interactions can be disturbed then the stability of the formed droplets might be affected in such a way that they might dissolve spontaneously.

We used simulations to investigate the molecular basis of αS monomeric aggregation into soluble oligomers resembling micro-LLPS. The WT protein demonstrated limited aggregation, suggesting a low inherent propensity for LLPS dictated by its primary sequence. IDPs, like αS, often share primary sequence characteristics associated with phase separation. Charged residues distributed with uncharged amino acids, resembling the ‘sticker and spacer’ model, contribute to this molecular grammar. This observation aligns with a general trend in IDPs (Choi et al., 2019; Martin et al., 2020; Choi et al., 2020). To assess αS LLPS propensity from its primary sequence, we calculated Shannon entropy (S) (Shannon, 1948, Equation 8 and Figure 8—source data 1), Kyte-Doolittle hydrophobicity (Kyte and Doolittle, 1982; Figure 8—source data 2), normalized, maximum of the sum of PLAAC log-likelihood ratios (NLLR) (Lancaster et al., 2014; Figure 8—source data 3), and LLPS propensity scores obtained from catGranules webserver (Bolognesi et al., 2016; Figure 8—source data 4).

where p_i is the probability of occurrence of a residue in a given sequence.

Comparative analysis with three datasets (Saar et al., 2021), namely LLPS+: a dataset of high propensity IDPs whose critical concentrations are 100 μM or below, LLPS-: a dataset of low propensity IDPs whose critical concentrations are greater than 100 μM, and PDB*: a dataset of folded proteins that do not undergo LLPS under normal conditions, revealed αS’s distinctive features (Figure 8—source data 5).

We note a significant difference in the Shannon entropy value of αS compared to proteins that do not undergo phase separation, as illustrated in Figure 8a. This deviation suggests a notable inclination of αS to undergo phase separation (Saar et al., 2021). Additionally, the hydrophobicity of αS (Figure 8b) is lower than that of the PDB* dataset, aligning more closely with the upper extremes of the LLPS- dataset. This indicates that while αS exhibits a tendency to undergo phase separation, the propensity should be low. Consistent with this, NLLR scores obtained from PLAAC and LLPS propensity scores (Figure 8c and d) reinforce this observation. These collective comparisons, coupled with simulations and experimental data on its critical concentration (Ray et al., 2020), conclusively establish that αS does not possess a high LLPS-forming propensity. Instead, this behavior is inherent to its primary structure. In hindsights, this analysis also justifies the requirements of environmental factors for enhancing the proclivity of αS for LLPS, as demonstrated in both our simulations and experimental findings (Ray et al., 2020; Sawner et al., 2021).

Figure 8

Download asset Open asset

Figure 8—source data 1

Shannon entropy (Shannon, 1948) for various datasets and α-synuclein (αS).

https://cdn.elifesciences.org/articles/95180/elife-95180-fig8-data1-v1.docx

Download elife-95180-fig8-data1-v1.docx

Figure 8—source data 2

Normalized Kyte-Doolittle hydrophobicity (Kyte and Doolittle, 1982) scores for various datasets and α-synuclein (αS).

https://cdn.elifesciences.org/articles/95180/elife-95180-fig8-data2-v1.docx

Download elife-95180-fig8-data2-v1.docx

Figure 8—source data 3

PLAAC normalized, maximum of the sum of PLAAC log-likelihood ratios (NLLR) (Lancaster et al., 2014) scores for various datasets and α-synuclein (αS).

https://cdn.elifesciences.org/articles/95180/elife-95180-fig8-data3-v1.docx

Download elife-95180-fig8-data3-v1.docx

Figure 8—source data 4

catGRANULE (Bolognesi et al., 2016) scores for various datasets and α-synuclein (αS).

https://cdn.elifesciences.org/articles/95180/elife-95180-fig8-data4-v1.docx

Download elife-95180-fig8-data4-v1.docx

Figure 8—source data 5

Comparison of primary sequence derived features for various datasets and α-synuclein (αS).

https://cdn.elifesciences.org/articles/95180/elife-95180-fig8-data5-v1.docx

Download elife-95180-fig8-data5-v1.docx

For characterizing αS’s aggregation phenomena, we calculated droplet surface tension under varied conditions. We observed that crowders minimally impacted surface tension, while salt increased it; however, both scenarios decreased the relative free energy of the system. Crowders achieved this via entropic means, whereas salt employed enthalpic means. Residue-residue interactions during droplet formation were consistent across environments, with crowder or salt enhancing these interactions. The aggregation pathway involved overall inter-chain interaction enhancement, specifically reducing intra-chain Nter-Cter and Nter-NAC interactions, leading to more extended protein conformations in droplets. The comparison with reported FRET observations (Ray et al., 2020) aligns well with the findings from our simulations, indicating that within the droplets, intra-chain NAC-NAC interactions have been supplanted by inter-chain NAC-NAC interactions. Droplet proteins displayed consistent orientation and ‘small-worldness’, a measure of inter-chain connectivity, remained consistent across diverse conditions. Thus, αS aggregates appeared invariant regarding their initial environment in terms of interactions and contacts.

Our study’s precision was notably influenced by the careful selection of a simulation force field. Despite the availability of modern force fields optimized for multi-chain simulations of IDPs (Dignon et al., 2018; Latham and Zhang, 2020; Regy et al., 2021; Zhang et al., 2022), we opted for Martini 3, an explicit water model, due to its emphasis on water’s role in aggregation and LLPS, as recently demonstrated in FUS LLPS (Mukherjee and Schäfer, 2023). Although newer models operate at a faster pace, Martini 3’s inclusion of explicit water enhances result accuracy. Additionally, Martini 3 provides a detailed amino acid description and allowing for encoding of protein secondary structures, unlike some newer models that represent amino acids as single beads. Our meticulous choice of the simulation model, combined with a comprehensive analysis, contributes to the accuracy and novelty of this study.

Recent studies have explored the aggregation and LLPS of biopolymers and polyelectrolytes in the presence of membranes, opening a promising avenue for αS research (Mondal and Cui, 2022; Liu et al., 2023a; Liu et al., 2023b). Given that under physiological conditions, αS assumes an oligomeric, membrane-bound form, investigating its interactions with membranes could hold therapeutic potential (Pineda and Burré, 2017).

Under physiological conditions, crowding effects emerge prominently. While crowders are commonly perceived to be inert, as has been considered in this investigation, the morphology, dimensions, and chemical interactions of crowding agents with αS in both dilute and dense phases may potentially exert considerable influence on its LLPS. Hence, a comprehensive understanding through systematic exploration is another avenue that warrants extensive investigation.

Although we exclusively focused on wild-type αS, familial mutations have been reported to exhibit a significantly higher propensity for aggregation (Ray et al., 2020). These mutations, involving minor alterations in the primary sequence, highlight the importance of understanding the molecular basis of this distinctive phenotype. Additionally, the observed stability of pre-formed αS droplets (Uversky et al., 2001) poses a challenge in treating PD. Reversing aggregation/LLPS and understanding associated pathways and mechanisms are crucial. Our study identifies key residues crucial for stable droplet formation, consistent across various environmental conditions.

The significance of the solvent in αS aggregation remains underexplored. Recent studies (Benayad et al., 2021; Mukherjee and Schäfer, 2023) underscore the pivotal role of water as a solvent in LLPS. It suggests that comprehending the solvent’s role, particularly water, is essential for attaining a deeper grasp of the thermodynamic and physical aspects of αS LLPS and aggregation. By delving into the solvent’s contribution, researchers can uncover additional factors influencing αS aggregation. Such insights hold the potential to advance our comprehension of protein aggregation phenomena, crucial for devising strategies to address diseases linked to protein misfolding and aggregation, notably PD. Future investigations focusing on elucidating the interplay between αS, solvent (especially water), and other environmental elements could yield valuable insights into the mechanisms underlying LLPS and aggregation. Ultimately, this could aid in the development of therapeutic interventions or preventive measures for Parkinson’s and related diseases.

All data are present within the manuscript. The sections of the trajectories used for analysis (1.5 - 2.0 µs) have been uploaded to zenodo (https://doi.org/10.5281/zenodo.10926367).Other relevant files related to the simulations have also been uploaded to the same repository.

1. 1. Wasim A
   2. Sneha M
   3. Jagannath M

   (2024) Zenodo

   Modulation of α-Synuclein Aggregation Amid Diverse Environmental Perturbation.

   https://doi.org/10.5281/zenodo.10926367

1. 1. Abraham MJ
   2. Murtola T
   3. Schulz R
   4. Páll S
   5. Smith JC
   6. Hess B
   7. Lindahl E

   (2015) GROMACS: high performance molecular simulations through multi-level parallelism from laptops to supercomputers

   SoftwareX 1–2:19–25.

   https://doi.org/10.1016/j.softx.2015.06.001

   • Google Scholar
2. 1. Adams GB
   2. O’Keeffe M
   3. Ruoff RS

   (1994) Van der waals surface areas and volumes of fullerenes

   The Journal of Physical Chemistry 98:9465–9469.

   https://doi.org/10.1021/j100089a018

   • Google Scholar
3. 1. Ahmed MC
   2. Skaanning LK
   3. Jussupow A
   4. Newcombe EA
   5. Kragelund BB
   6. Camilloni C
   7. Langkilde AE
   8. Lindorff-Larsen K

   (2021) Refinement of α-synuclein ensembles against SAXS data: comparison of force fields and methods

   Frontiers in Molecular Biosciences 8:654333.

   https://doi.org/10.3389/fmolb.2021.654333

   • PubMed
   • Google Scholar
4. 1. Apetri MM
   2. Maiti NC
   3. Zagorski MG
   4. Carey PR
   5. Anderson VE

   (2006) Secondary structure of alpha-synuclein oligomers: characterization by raman and atomic force microscopy

   Journal of Molecular Biology 355:63–71.

   https://doi.org/10.1016/j.jmb.2005.10.071

   • PubMed
   • Google Scholar
5. 1. Bah A
   2. Forman-Kay JD

   (2016) Modulation of intrinsically disordered protein function by post-translational modifications

   The Journal of Biological Chemistry 291:6696–6705.

   https://doi.org/10.1074/jbc.R115.695056

   • PubMed
   • Google Scholar
6. 1. Banani SF
   2. Lee HO
   3. Hyman AA
   4. Rosen MK

   (2017) Biomolecular condensates: organizers of cellular biochemistry

   Nature Reviews. Molecular Cell Biology 18:285–298.

   https://doi.org/10.1038/nrm.2017.7

   • PubMed
   • Google Scholar
7. 1. Bari KJ
   2. Prakashchand DD

   (2021) Fundamental challenges and outlook in simulating liquid-liquid phase separation of intrinsically disordered proteins

   The Journal of Physical Chemistry Letters 12:1644–1656.

   https://doi.org/10.1021/acs.jpclett.0c03404

   • PubMed
   • Google Scholar
8. 1. Barrat A
   2. Weigt M

   (2000) On the properties of small-world network models

   The European Physical Journal B 13:547–560.

   https://doi.org/10.1007/s100510050067

   • Google Scholar
9. 1. Benayad Z
   2. von Bülow S
   3. Stelzl LS
   4. Hummer G

   (2021) Simulation of FUS protein condensates with an adapted coarse-grained model

   Journal of Chemical Theory and Computation 17:525–537.

   https://doi.org/10.1021/acs.jctc.0c01064

   • PubMed
   • Google Scholar
10. 1. Best RB
    2. Zheng W
    3. Mittal J

    (2014) Balanced protein-water interactions improve properties of disordered proteins and non-specific protein association

    Journal of Chemical Theory and Computation 10:5113–5124.

    https://doi.org/10.1021/ct500569b

    • PubMed
    • Google Scholar
11. 1. Bolognesi B
    2. Lorenzo Gotor N
    3. Dhar R
    4. Cirillo D
    5. Baldrighi M
    6. Tartaglia GG
    7. Lehner B

    (2016) A concentration-dependent liquid phase separation can cause toxicity upon increased protein expression

    Cell Reports 16:222–231.

    https://doi.org/10.1016/j.celrep.2016.05.076

    • PubMed
    • Google Scholar
12. 1. Chen SW
    2. Drakulic S
    3. Deas E
    4. Ouberai M
    5. Aprile FA
    6. Arranz R
    7. Ness S
    8. Roodveldt C
    9. Guilliams T
    10. De-Genst EJ
    11. Klenerman D
    12. Wood NW
    13. Knowles TPJ
    14. Alfonso C
    15. Rivas G
    16. Abramov AY
    17. Valpuesta JM
    18. Dobson CM
    19. Cremades N

    (2015) Structural characterization of toxic oligomers that are kinetically trapped during α-synuclein fibril formation

    PNAS 112:E1994–E2003.

    https://doi.org/10.1073/pnas.1421204112

    • PubMed
    • Google Scholar
13. 1. Choi JM
    2. Dar F
    3. Pappu RV

    (2019) LASSI: a lattice model for simulating phase transitions of multivalent proteins

    PLOS Computational Biology 15:e1007028.

    https://doi.org/10.1371/journal.pcbi.1007028

    • PubMed
    • Google Scholar
14. 1. Choi JM
    2. Holehouse AS
    3. Pappu RV

    (2020) Physical principles underlying the complex biology of intracellular phase transitions

    Annual Review of Biophysics 49:107–133.

    https://doi.org/10.1146/annurev-biophys-121219-081629

    • PubMed
    • Google Scholar
15. Book

    1. Cremades N
    2. Chen SW
    3. Dobson CM

    (2017) Chapter three - structural characteristics of -synuclein oligomers

    In: Sandal M, editors. Early Stage Protein Misfolding and Amyloid Aggregation Volume 329 of International Review of Cell and Molecular Biology. Academic Press. pp. 79–143.

    https://doi.org/10.1016/bs.ircmb.2016.08.010

    • Google Scholar
16. 1. Deeds EJ
    2. Ashenberg O
    3. Gerardin J
    4. Shakhnovich EI

    (2007) Robust protein protein interactions in crowded cellular environments

    PNAS 104:14952–14957.

    https://doi.org/10.1073/pnas.0702766104

    • PubMed
    • Google Scholar
17. 1. Dignon GL
    2. Zheng W
    3. Kim YC
    4. Best RB
    5. Mittal J

    (2018) Sequence determinants of protein phase behavior from a coarse-grained model

    PLOS Computational Biology 14:e1005941.

    https://doi.org/10.1371/journal.pcbi.1005941

    • PubMed
    • Google Scholar
18. 1. Ellis RJ
    2. Minton AP

    (2006) Protein aggregation in crowded environments

    Biological Chemistry 387:485–497.

    https://doi.org/10.1515/BC.2006.064

    • PubMed
    • Google Scholar
19. 1. Farag M
    2. Cohen SR
    3. Borcherds WM
    4. Bremer A
    5. Mittag T
    6. Pappu RV

    (2022) Condensates formed by prion-like low-complexity domains have small-world network structures and interfaces defined by expanded conformations

    Nature Communications 13:7722.

    https://doi.org/10.1038/s41467-022-35370-7

    • PubMed
    • Google Scholar
20. 1. Forman-Kay JD
    2. Mittag T

    (2013) From sequence and forces to structure, function, and evolution of intrinsically disordered proteins

    Structure 21:1492–1499.

    https://doi.org/10.1016/j.str.2013.08.001

    • PubMed
    • Google Scholar
21. 1. Fuxreiter M
    2. Tompa P

    (2012) Fuzzy complexes: a more stochastic view of protein function

    Fuzziness: Structural Disorder in Protein Complexes 01:1–14.

    https://doi.org/10.1007/978-1-4614-0659-4_1

    • Google Scholar
22. Book

    1. Gowers R
    2. Linke M
    3. Barnoud J
    4. Reddy T
    5. Melo M
    6. Seyler S
    7. Domański J
    8. Dotson D
    9. Buchoux S
    10. Kenney I
    11. Beckstein O

    (2016) MDAnalysis: A python package for the rapid analysis of molecular dynamics simulations

    In: Linke M, Reddy T, editors. Proceedings of the 15th Python in Science Conference. SciPy Austin. pp. 1–105.

    https://doi.org/10.25080/Majora-629e541a-00e

    • Google Scholar
23. 1. Gui X
    2. Feng S
    3. Li Z
    4. Li Y
    5. Reif B
    6. Shi B
    7. Niu Z

    (2023) Liquid-liquid phase separation of amyloid-β oligomers modulates amyloid fibrils formation

    The Journal of Biological Chemistry 299:102926.

    https://doi.org/10.1016/j.jbc.2023.102926

    • PubMed
    • Google Scholar
24. 1. Hong DP
    2. Han S
    3. Fink AL
    4. Uversky VN

    (2011) Characterization of the non-fibrillar α-synuclein oligomers

    Protein and Peptide Letters 18:230–240.

    https://doi.org/10.2174/092986611794578332

    • PubMed
    • Google Scholar
25. 1. Horvath I
    2. Kumar R
    3. Wittung-Stafshede P

    (2021) Macromolecular crowding modulates α-synuclein amyloid fiber growth

    Biophysical Journal 120:3374–3381.

    https://doi.org/10.1016/j.bpj.2021.06.032

    • PubMed
    • Google Scholar
26. 1. Huang J
    2. Rauscher S
    3. Nawrocki G
    4. Ran T
    5. Feig M
    6. de Groot BL
    7. Grubmüller H
    8. MacKerell AD

    (2017) CHARMM36m: an improved force field for folded and intrinsically disordered proteins

    Nature Methods 14:71–73.

    https://doi.org/10.1038/nmeth.4067

    • PubMed
    • Google Scholar
27. 1. Humphries MD
    2. Gurney K

    (2008) Network “small-world-ness”: a quantitative method for determining canonical network equivalence

    PLOS ONE 3:e0002051.

    https://doi.org/10.1371/journal.pone.0002051

    • PubMed
    • Google Scholar
28. 1. Hunter JD

    (2007) Matplotlib: a 2d graphics environment

    Computing in Science & Engineering 9:90–95.

    https://doi.org/10.1109/MCSE.2007.55

    • Google Scholar
29. 1. Hyman AA
    2. Weber CA
    3. Jülicher F

    (2014) Liquid-liquid phase separation in biology

    Annual Review of Cell and Developmental Biology 30:39–58.

    https://doi.org/10.1146/annurev-cellbio-100913-013325

    • PubMed
    • Google Scholar
30. Website

    1. Inkscape

    (2024) Inkscape Project

    Accessed May 16, 2024.

    https://inkscape.org/
31. 1. Joosten RP
    2. te Beek TAH
    3. Krieger E
    4. Hekkelman ML
    5. Hooft RWW
    6. Schneider R
    7. Sander C
    8. Vriend G

    (2011) A series of PDB related databases for everyday needs

    Nucleic Acids Research 39:D411–D419.

    https://doi.org/10.1093/nar/gkq1105

    • PubMed
    • Google Scholar
32. 1. Joseph JA
    2. Reinhardt A
    3. Aguirre A
    4. Chew PY
    5. Russell KO
    6. Espinosa JR
    7. Garaizar A
    8. Collepardo-Guevara R

    (2021) Physics-driven coarse-grained model for biomolecular phase separation with near-quantitative accuracy

    Nature Computational Science 1:732–743.

    https://doi.org/10.1038/s43588-021-00155-3

    • PubMed
    • Google Scholar
33. 1. Kabsch W
    2. Sander C

    (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features

    Biopolymers 22:2577–2637.

    https://doi.org/10.1002/bip.360221211

    • PubMed
    • Google Scholar
34. Book

    1. Kluyver T
    2. Ragan-Kelley B
    3. Pérez F
    4. Granger B
    5. Bussonnier M
    6. Frederic J
    7. Kelley K
    8. Hamrick J
    9. Grout J
    10. Corlay S
    11. Ivanov P
    12. Avila D
    13. Abdalla S
    14. Willing C

    (2016)

    Jupyter notebooks – a publishing format for reproducible computational workflows

    In: Loizides F, Schmidt B, editors. Positioning and Power in Academic Publishing: Players, Agents and Agendas. IOS Press. pp. 87–90.

    • Google Scholar
35. 1. Kyte J
    2. Doolittle RF

    (1982) A simple method for displaying the hydropathic character of A protein

    Journal of Molecular Biology 157:105–132.

    https://doi.org/10.1016/0022-2836(82)90515-0

    • PubMed
    • Google Scholar
36. 1. Lancaster AK
    2. Nutter-Upham A
    3. Lindquist S
    4. King OD

    (2014) PLAAC: a web and command-line application to identify proteins with prion-like amino acid composition

    Bioinformatics 30:2501–2502.

    https://doi.org/10.1093/bioinformatics/btu310

    • PubMed
    • Google Scholar
37. 1. Latham AP
    2. Zhang B

    (2020) Maximum entropy optimized force field for intrinsically disordered proteins

    Journal of Chemical Theory and Computation 16:773–781.

    https://doi.org/10.1021/acs.jctc.9b00932

    • PubMed
    • Google Scholar
38. 1. Li E
    2. Placone J
    3. Merzlyakov M
    4. Hristova K

    (2008) Quantitative measurements of protein interactions in a crowded cellular environment

    Analytical Chemistry 80:5976–5985.

    https://doi.org/10.1021/ac800616u

    • PubMed
    • Google Scholar
39. 1. Liu Z
    2. Yethiraj A
    3. Cui Q

    (2023a) Selective and sensitive surface condensation driven by coupled phase behaviors of membrane and biopolymers

    Biophysical Journal 122:207a–208a.

    https://doi.org/10.1016/j.bpj.2022.11.1247

    • Google Scholar
40. 1. Liu Z
    2. Yethiraj A
    3. Cui Q

    (2023b) Sensitive and selective polymer condensation at membrane surface driven by positive co-operativity

    PNAS 120:e2212516120.

    https://doi.org/10.1073/pnas.2212516120

    • Google Scholar
41. 1. Martin EW
    2. Holehouse AS
    3. Peran I
    4. Farag M
    5. Incicco JJ
    6. Bremer A
    7. Grace CR
    8. Soranno A
    9. Pappu RV
    10. Mittag T

    (2020) Valence and patterning of aromatic residues determine the phase behavior of prion-like domains

    Science 367:694–699.

    https://doi.org/10.1126/science.aaw8653

    • PubMed
    • Google Scholar
42. 1. Menon S
    2. Mondal J

    (2022) Small molecule modulatesΑ-synuclein conformation and its oligomerization via entropy expansion

    Biophysics 01:e3005.

    https://doi.org/10.1101/2022.10.20.513005

    • PubMed
    • Google Scholar
43. 1. Menon S
    2. Mondal J

    (2023) Conformational plasticity in α-synuclein and how crowded environment modulates it

    The Journal of Physical Chemistry. B 127:4032–4049.

    https://doi.org/10.1021/acs.jpcb.3c00982

    • PubMed
    • Google Scholar
44. 1. Michaud-Agrawal N
    2. Denning EJ
    3. Woolf TB
    4. Beckstein O

    (2011) MDAnalysis: a toolkit for the analysis of molecular dynamics simulations

    Journal of Computational Chemistry 32:2319–2327.

    https://doi.org/10.1002/jcc.21787

    • PubMed
    • Google Scholar
45. 1. Mondal S
    2. Cui Q

    (2022) Coacervation of poly-electrolytes in the presence of lipid bilayers: mutual alteration of structure and morphology

    Chemical Science 13:7933–7946.

    https://doi.org/10.1039/d2sc02013k

    • PubMed
    • Google Scholar
46. 1. Mukherjee S
    2. Sakunthala A
    3. Gadhe L
    4. Poudyal M
    5. Sawner AS
    6. Kadu P
    7. Maji SK

    (2023) Liquid-liquid phase separation of α-synuclein: a new mechanistic insight for α-synuclein aggregation associated with parkinson’s disease pathogenesis

    Journal of Molecular Biology 435:167713.

    https://doi.org/10.1016/j.jmb.2022.167713

    • PubMed
    • Google Scholar
47. 1. Mukherjee S
    2. Schäfer LV

    (2023) Thermodynamic forces from protein and water govern condensate formation of an intrinsically disordered protein domain

    Nature Communications 14:5892.

    https://doi.org/10.1038/s41467-023-41586-y

    • PubMed
    • Google Scholar
48. 1. Munishkina LA
    2. Ahmad A
    3. Fink AL
    4. Uversky VN

    (2008) Guiding protein aggregation with macromolecular crowding

    Biochemistry 47:8993–9006.

    https://doi.org/10.1021/bi8008399

    • PubMed
    • Google Scholar
49. 1. Nguyen HT
    2. Hori N
    3. Thirumalai D

    (2022) Condensates in RNA repeat sequences are heterogeneously organized and exhibit reptation dynamics

    Nature Chemistry 14:775–785.

    https://doi.org/10.1038/s41557-022-00934-z

    • PubMed
    • Google Scholar
50. 1. Piana S
    2. Robustelli P
    3. Tan D
    4. Chen S
    5. Shaw DE

    (2020) Development of a force field for the simulation of single-chain proteins and protein-protein complexes

    Journal of Chemical Theory and Computation 16:2494–2507.

    https://doi.org/10.1021/acs.jctc.9b00251

    • PubMed
    • Google Scholar
51. 1. Pineda A
    2. Burré J

    (2017) Modulating membrane binding of α-synuclein as a therapeutic strategy

    PNAS 114:1223–1225.

    https://doi.org/10.1073/pnas.1620159114

    • PubMed
    • Google Scholar
52. 1. Ray S
    2. Singh N
    3. Kumar R
    4. Patel K
    5. Pandey S
    6. Datta D
    7. Mahato J
    8. Panigrahi R
    9. Navalkar A
    10. Mehra S
    11. Gadhe L
    12. Chatterjee D
    13. Sawner AS
    14. Maiti S
    15. Bhatia S
    16. Gerez JA
    17. Chowdhury A
    18. Kumar A
    19. Padinhateeri R
    20. Riek R
    21. Krishnamoorthy G
    22. Maji SK

    (2020) α-Synuclein aggregation nucleates through liquid-liquid phase separation

    Nature Chemistry 12:705–716.

    https://doi.org/10.1038/s41557-020-0465-9

    • PubMed
    • Google Scholar
53. 1. Regy RM
    2. Thompson J
    3. Kim YC
    4. Mittal J

    (2021) Improved coarse-grained model for studying sequence dependent phase separation of disordered proteins

    Protein Science 30:1371–1379.

    https://doi.org/10.1002/pro.4094

    • PubMed
    • Google Scholar
54. 1. Rényi E

    (1959) On random graphs

    Publicationes Mathematicae Debrecen 6:290–297.

    https://doi.org/10.5486/PMD.1959.6.3-4.12

    • Google Scholar
55. 1. Robustelli P
    2. Piana S
    3. Shaw DE

    (2018) Developing a molecular dynamics force field for both folded and disordered protein states

    PNAS 115:E4758–E4766.

    https://doi.org/10.1073/pnas.1800690115

    • PubMed
    • Google Scholar
56. 1. Rodríguez LC
    2. Foressi NN
    3. Celej MS

    (2023) Modulation of α-synuclein phase separation by biomolecules

    Biochimica et Biophysica Acta - Proteins and Proteomics 1871:140885.

    https://doi.org/10.1016/j.bbapap.2022.140885

    • Google Scholar
57. 1. Saar KL
    2. Morgunov AS
    3. Qi R
    4. Arter WE
    5. Krainer G
    6. Lee AA
    7. Knowles TPJ

    (2021) Learning the molecular grammar of protein condensates from sequence determinants and embeddings

    PNAS 118:e2019053118.

    https://doi.org/10.1073/pnas.2019053118

    • PubMed
    • Google Scholar
58. 1. Sawner AS
    2. Ray S
    3. Yadav P
    4. Mukherjee S
    5. Panigrahi R
    6. Poudyal M
    7. Patel K
    8. Ghosh D
    9. Kummerant E
    10. Kumar A

    (2021) Modulating -synuclein liquid–liquid phase separation: Published as part of the biochemistry virtual special issue “protein condensates

    Biochemistry 60:3676–3696.

    https://doi.org/10.1021/acs.biochem.1c00434

    • Google Scholar
59. Software

    1. Schrödinger LLC

    (2015a) The axpymol molecular graphics plugin for microsoft powerpoint, version 1.8

    PyMOL.

    https://www.pymol.org/ax.html?
60. Software

    1. Schrödinger LLC

    (2015b) The jymol molecular graphics development component version, version 1.8

    PyMOL.

    https://www.pymol.org/
61. Software

    1. Schrödinger LLC

    (2015c) The pymol molecular graphics system, version 1.8

    PyMOL.

    https://pymol.org/
62. 1. Shannon CE

    (1948) A mathematical theory of communication

    Bell System Technical Journal 27:379–423.

    https://doi.org/10.1002/j.1538-7305.1948.tb01338.x

    • Google Scholar
63. 1. Shin Y
    2. Brangwynne CP

    (2017) Liquid phase condensation in cell physiology and disease

    Science 357:eaaf4382.

    https://doi.org/10.1126/science.aaf4382

    • PubMed
    • Google Scholar
64. 1. Shu T
    2. Alberts J
    3. Sang D
    4. Tsao S
    5. Palhares Viana M
    6. Heidenreich M
    7. Levy E
    8. Holt LJ

    (2021) Liquid-liquid phase separation is influenced by the active, crowded cytoplasm

    Biophysical Journal 120:31a–32a.

    https://doi.org/10.1016/j.bpj.2020.11.445

    • Google Scholar
65. 1. Souza PCT
    2. Alessandri R
    3. Barnoud J
    4. Thallmair S
    5. Faustino I
    6. Grünewald F
    7. Patmanidis I
    8. Abdizadeh H
    9. Bruininks BMH
    10. Wassenaar TA
    11. Kroon PC
    12. Melcr J
    13. Nieto V
    14. Corradi V
    15. Khan HM
    16. Domański J
    17. Javanainen M
    18. Martinez-Seara H
    19. Reuter N
    20. Best RB
    21. Vattulainen I
    22. Monticelli L
    23. Periole X
    24. Tieleman DP
    25. de Vries AH
    26. Marrink SJ

    (2021) Martini 3: a general purpose force field for coarse-grained molecular dynamics

    Nature Methods 18:382–388.

    https://doi.org/10.1038/s41592-021-01098-3

    • PubMed
    • Google Scholar
66. 1. Tesei G
    2. Lindorff-Larsen K

    (2022) Improved predictions of phase behaviour of intrinsically disordered proteins by tuning the interaction range

    Open Research Europe 2:94.

    https://doi.org/10.12688/openreseurope.14967.2

    • PubMed
    • Google Scholar
67. 1. Thomasen FE
    2. Pesce F
    3. Roesgaard MA
    4. Tesei G
    5. Lindorff-Larsen K

    (2022) Improving martini 3 for disordered and multidomain proteins

    Journal of Chemical Theory and Computation 18:2033–2041.

    https://doi.org/10.1021/acs.jctc.1c01042

    • PubMed
    • Google Scholar
68. 1. Touw WG
    2. Baakman C
    3. Black J
    4. te Beek TAH
    5. Krieger E
    6. Joosten RP
    7. Vriend G

    (2015) A series of PDB-related databanks for everyday needs

    Nucleic Acids Research 43:D364–D368.

    https://doi.org/10.1093/nar/gku1028

    • PubMed
    • Google Scholar
69. 1. Ubbiali D
    2. Fratini M
    3. Piersimoni L
    4. Ihling CH
    5. Kipping M
    6. Heilmann I
    7. Iacobucci C
    8. Sinz A

    (2022) Direct observation of “elongated” conformational states in α-synuclein upon liquid-liquid phase separation

    Angewandte Chemie 61:e202205726.

    https://doi.org/10.1002/anie.202205726

    • PubMed
    • Google Scholar
70. 1. Uversky VN
    2. Li J
    3. Fink AL

    (2001) Evidence for a partially folded intermediate in alpha-synuclein fibril formation

    The Journal of Biological Chemistry 276:10737–10744.

    https://doi.org/10.1074/jbc.M010907200

    • PubMed
    • Google Scholar
71. 1. Uversky VN
    2. M Cooper E
    3. Bower KS
    4. Li J
    5. Fink AL

    (2002) Accelerated alpha-synuclein fibrillation in crowded milieu

    FEBS Letters 515:99–103.

    https://doi.org/10.1016/s0014-5793(02)02446-8

    • PubMed
    • Google Scholar
72. 1. Uversky VN
    2. Oldfield CJ
    3. Dunker AK

    (2008) Intrinsically disordered proteins in human diseases: introducing the D2 concept

    Annual Review of Biophysics 37:215–246.

    https://doi.org/10.1146/annurev.biophys.37.032807.125924

    • PubMed
    • Google Scholar
73. 1. Van Der Spoel D
    2. Lindahl E
    3. Hess B
    4. Groenhof G
    5. Mark AE
    6. Berendsen HJC

    (2005) GROMACS: fast, flexible, and free

    Journal of Computational Chemistry 26:1701–1718.

    https://doi.org/10.1002/jcc.20291

    • PubMed
    • Google Scholar
74. Book

    1. Van Rossum G
    2. Drake FL

    (2009)

    Python 3 Reference Manual

    CreateSpace.

    • Google Scholar
75. 1. Watts DJ
    2. Strogatz SH

    (1998) Collective dynamics of “small-world” networks

    Nature 393:440–442.

    https://doi.org/10.1038/30918

    • PubMed
    • Google Scholar
76. 1. Xing Y
    2. Nandakumar A
    3. Kakinen A
    4. Sun Y
    5. Davis TP
    6. Ke PC
    7. Ding F

    (2021) Amyloid aggregation under the lens of liquid-liquid phase separation

    The Journal of Physical Chemistry Letters 12:368–378.

    https://doi.org/10.1021/acs.jpclett.0c02567

    • PubMed
    • Google Scholar
77. 1. Zerze GH

    (2024) Optimizing the martini 3 force field reveals the effects of the intricate balance between protein-water interaction strength and salt concentration on biomolecular condensate formation

    Journal of Chemical Theory and Computation 20:1646–1655.

    https://doi.org/10.1021/acs.jctc.2c01273

    • PubMed
    • Google Scholar
78. 1. Zhang Y
    2. Liu X
    3. Chen J

    (2022) Toward accurate coarse-grained simulations of disordered proteins and their dynamic interactions

    Journal of Chemical Information and Modeling 62:4523–4536.

    https://doi.org/10.1021/acs.jcim.2c00974

    • PubMed
    • Google Scholar
79. 1. Zhou HX

    (2013) Influence of crowded cellular environments on protein folding, binding, and oligomerization: biological consequences and potentials of atomistic modeling

    FEBS Letters 587:1053–1061.

    https://doi.org/10.1016/j.febslet.2013.01.064

    • PubMed
    • Google Scholar

Author details

1. Abdul Wasim

   Tata Institute of Fundamental Research, Hyderabad, India

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Methodology, Writing - original draft, Writing - review and editing

   Competing interests

   No competing interests declared

2. Sneha Menon

   Tata Institute of Fundamental Research, Hyderabad, India

   Contribution

   Formal analysis, Writing - original draft

   Competing interests

   No competing interests declared

3. Jagannath Mondal

   Tata Institute of Fundamental Research, Hyderabad, India

   Contribution

   Conceptualization, Supervision, Funding acquisition, Validation, Writing - original draft, Project administration, Writing - review and editing

   For correspondence

   jmondal@tifrh.res.in

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1090-5199

• 465

  views

• 27

  downloads

• 0

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.