Mutation/metal deficiency in the "electrostatic loop" enhanced aggregation process in apo/holo SOD1 variants: implications for ALS diseases

Abstract

Despite the many mechanisms it has created to prevent unfolding and aggregation of proteins, many diseases are caused by abnormal folding of proteins, which are called misfolding diseases. During this process, proteins undergo structural changes and become stable, insoluble beta-sheet aggregates called amyloid fibrils. Mutations/disruptions in metal ion homeostasis in the ALS-associated metalloenzyme superoxide dismutase (SOD1) reduce conformational stability, consistent with the protein aggregation hypothesis for neurodegenerative diseases. However, the exact mechanism of involvement is not well understood. Hence, to understand the role of mutation/ metal deficiency in SOD1 misfolding and aggregation, we investigated the effects of apo/holo SOD1 variants on structural properties using biophysical/experimental techniques. The MD results support the idea that the mutation/metal deficiency can lead to a change in conformation. The increased content of β-sheet structures in apo/holo SOD1 variants can be attributed to the aggregation tendency, which was confirmed by FTIR spectroscopy and dictionary of secondary structure in proteins (DSSP) results. Thermodynamic studies of GdnHCl showed that metal deficiency/mutation/intramolecular S–S reduction together are required to initiate misfolding/aggregation of SOD1. The results showed that apo/holo SOD1 variants under destabilizing conditions induced amyloid aggregates at physiological pH, which were detected by ThT/ANS fluorescence, as well as further confirmation of amyloid/amorphous species by TEM. This study confirms that mutations in the electrostatic loop of SOD1 lead to structural abnormalities, including changes in hydrophobicity, reduced disulfide bonds, and an increased propensity for protein denaturation. This process facilitates the formation of amyloid/amorphous aggregates ALS-associated.

Introduction

Protein folding has been one of the most basic biological processes ensuing protein synthesis, which has been recently of interest among academics [1]. As a complex and spontaneous physicochemical process, its involves a series of incompletely folded intermediate stages toward a functionally folded form [2]. In contrast, protein misfolding can be the occasion of diseases caused by pathogens or loss of protein functions due to aggregated misfolded proteins. Accordingly, a collection of illnesses, known as protein misfolding disorders or PMDs, have been defined by the aggregation of incorrectly folded proteins [3]. In this context, amyloidogenic proteins are associated with PMDs as the conditions when there are misfolded monomeric proteins that form insoluble amyloid fibrils on their own [4]. Unstable misaggregated species thus oligomerize, and then act as aggregates creating more misfolding and forming fibrillar structures [5]. Toxic protein aggregates have been further connected to some human illnesses, including neurodegenerative diseases (NGDs) such as amyotrophic lateral sclerosis (ALS), wherein the motor neurons of the motor cortex, brainstem, and spinal cord are disturbed by this neurological condition, and finally malfunction and death occur [6]. Most cases with ALS are sporadic (sALS) and no more than 5–10% of the cases have been reported to be familial (fALS). In this domain, the point mutations taking place in the antioxidant enzyme superoxide dismutase 1 (SOD1) have been introduced as the foremost genetic basis for fALS [7]. Of note, mature metallated-SOD1 is a stable protein because it is dimeric and characterized by copper-zinc (Cu–Zn) binding as well as intramolecular disulfide (S–S) links. The development of immature forms of SOD1, which seem to be susceptible to aggregation and induce ALS accordingly arise out of absence or modification of each post-translational process or disruptions by some mutations [8]. No correlation between SOD1 activity and disease progression additionally implies that loss of function in SOD1 is not to blame for ALS [9]. In this way, the pathophysiology of mutant SOD1-associated motor neuron disorders in view of the loss-of-function explanation has been replaced by the theory of toxic gain of function [10]. As reported, there have been more than 220 SOD1 mutations in cases with ALS [11]. Much effort has been further made to characterize the biophysical properties of multiple SOD1 mutations associated with ALS. In spite of this, the molecular processes underlying the mutations leading to fALS are not properly understood [12]. The mutations tied to ALS-associated SOD1 aggregation can disrupt dynamic coupling between monomers [13], cause dimer dissociation and/or apo-monomer formation, and ultimately form fibrils formation [14]. On the other hand, each mutant is likely to form fibrils with distinct shapes, and such variations are attributed to some modifications in local residue structures and protein dynamics [15]. The mutations in SOD1, linked to ALS, seem to upset the stable structure of enzymes. This is accompanied by reduced metal ion concentration, fewer intramolecular S–S bonds, loss of post-translational modifications, less charge, disruption in surface hydrogen (H) bonding network, and then accelerated development of mutant SOD1 aggregates that bear a resemblance to amyloids [16,17,18,19,20,21,22,23,24,25,26]. As acknowledged, one of the core causes of SOD1 aggregation might be metal deficiency or some alterations in hydrophobicity. However, recent studies have focused on the ways aberrant metal-binding is likely to influence SOD1 aggregation [27, 28]. Notably, it has been suggested that abnormal metal-binding and demetallization increase SOD1 misfolding and aggregation, indicating the potential role of metal binding in SOD1 pathological aggregation [29]. Some cases with a heterozygous point mutation in exon 5 of SOD1 (c.377T > C), resulting in the substitution of amino acid leucine for serine at codon 126 (p.L126S) on a highly conserved position, have also presented significant progression [30]. Among the others, L126S mutation has negative effects on SOD1 stability and function and even makes it destabilized. In this regard, ALS pathophysiology has conventionally been linked to electrostatic loop destabilization by disease-inducing mutations [31]. Moreover, aggregation-prone mutants are likely to exhibit hydrophobicity toward polar uncharged substituents at the rate of 48%, thereby drawing much attention to the impact of modified hydrophobicity on aggregation. Of note, SOD1 mutations are more disposed to aggregation due to some changes in hydrophobicity, loss of native core packing, and a set of destabilizing states. ALS-associated mutations also significantly reduce the interaction energy of contact sites between electrostatic loop and remaining SOD1 structure [32]. Considering the folding and aggregation mechanism of SOD1, knowing more about the dynamic role of Zn and Cu in such incidents is of utmost importance. To reflect on their roles, there was a need to access the data on the dynamic properties of SOD1 mutants in protein aggregation or activity at the molecular level. In fact, a series of experimental approaches are now being practiced to improve dynamic properties. As a first option, molecular dynamics (MD), a computer-based simulation method, has been proposed as an alternative to explore conformational changes engendered by mutations and broaden more atomic-level knowledge regarding the structure and dynamics of proteins plus their biological functions. Therefore, simulations as the most powerful tools help figure out disease-related mutations and their consequences on the structure of proteins. Against this background, the present study was an attempt to shed light on the structural features of the apo/holo forms of the wild-type (WT)-SOD1 and L126S mutants in an electrostatic loop, using computational and experimental techniques in order to characterize some physicochemical properties, viz., alterations in hydrophobicity, mutation/metal deficiency, and then their correlation with molecular-level protein aggregation. In this manner, this study aimed to facilitate better understanding of the mechanism underlying amyloid aggregation associated with L126S mutations forming metal-binding sites, and consequently explore its impact on ALS.

Materials and methods

Computational techniques

Structural changes and activity stability using bioinformatics

The ways amino acid substitution affects protein structure and function might be investigated through various computational methods. To forecast activity stability in the mutant protein, the Predicts was used [33]. The effect of mutation on human (h)SOD1 stability and structure was then evaluated by the free energy change (ΔΔG) calculations for the I-Mutant, i-Stable, and DUET servers, which could report ΔΔG as positive and negative values for stabilizing and destabilizing mutations, respectively [34].

Molecular dynamics (MD)

With regard to the research design, the Protein Data Bank (PDB) affected to the Research Collaboratory for Structural Bioinformatics (PDB ID: 2C9V), with the 1.07 Å resolution, was utilized in order to obtain the primary structure of WT-SOD1 for MD-based simulations [35], as they could determine the effects of mutagenesis on protein structure and describe destabilization and stabilization of the mechanisms of interest. The simulations of the structures of the apo/holo-SOD1 variants were accordingly performed via the GROMACS 4.6.5 software package with the GROMOSE96 54A7 force field [36]. For this purpose, the models were initially put at the central part of a dodecahedral box, and then solvated with a simple point charge (SPC) water model. The chloride (Cl^¯) or sodium (Na⁺) ions were further employed to neutralize each system. The Cu and Zn metal ions were then position-restrained to their relevant binding sites, but just in the holo form [37]. The system energy was also minimized using the Method of Steepest Descent. Afterward, the equilibrations of the systems were done under constant temperature, constant volume (NVT) up to 100 ps at 310 K with restraint forces of 1000 kJ/mol, followed by 100 ps under constant temperature, constant pressure (NPT) of 1 bar with restraint forces of 1000 kJ/mol. Furthermore, the Leapfrog algorithm with the time step of 2fs was operated to integrate Newton’s equation in the MD-based simulation and collect the data every 10ps [38]. On the other hand, the Library of Integrated Network-Based Cellular Signatures (LINCS) Program was applied to maintain the covalent bonds of proteins during fixed binding. The Particle-Mesh Ewald (PME) technique [39] was then employed for treating electrostatic interactions and MD-based simulations of the equilibrated system for 150 ns, in which each atom could change its positions without restrictions.

Experiment

Materials

The agarose resins, nickel-nitrilotriacetic acid (Ni–NTA) and 1-anilinonaphthalene-8-sulfonate (ANS) and were originally purchased from the Merck Group Inc. and the QIAGEN (Germany), respectively. In addition, the protein ladder (10–250 kDa), boric acid (H₃BO₃), trisaminomethane/hydrochloride (Tris/HCl), dithiothreitol (DTT), phenylmethylsulfonyl fluoride (PMSF), sodium chloride (NaCl), Congo red (CR), thioflavin-T (ThT), sodium acetate (NaOAc), kanamycin, isopropyl-b-D-thiogalactopyranoside (IPTG), tryptone, yeast extract, and agarose, were acquired from the Bio Basic Inc. (Canada). Besides, the dialysis tubing cellulose membrane with the 12400 Da cut-off and the sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) were procured from the Sigma-Aldrich and the Merck Group Inc. (Germany), in that order. Moreover, the plasmid extraction kit, Dpn I, and the Pfu DNA polymerase were bought from the Bioneer Corp. (South Korea) and the Thermo Fisher Scientific Inc. (the USA), respectively.

Site-directed mutagenesis and WT-SOD1 and mutant expression and purification

The recombinant vector, pET-28a ( +), with the SOD1 gene (pET28a-SOD1) as the template, was employed here. Then, the Quick-change PCR site-directed mutagenesis was carried out to form the single-point mutation L126S. In view of that, the point mutation was developed through designing two mutagenesis primers for L126S, the forward primer (5ʹ AAAAAGCAGATGACTCGGGCAAAGGTGGAAATG 3ʹ) and the reverse one (5ʹ TCCACCTTTGCCCGAGTCATCTGCTTTTTCATG 3ʹ) with the mutation sites (as highlighted). As well, the apo/holo-WT and mutant proteins were induced in the E. coli-BL21 (DE3) with 0.8 mM isopropyl β-D-1-thiogalactopyranoside (IPTG), 60 μM zinc sulfate (ZnSO₄), 7 mM lactose, and 200 μM copper sulfate (CuSO₄) (just for the holo-SOD1), lasting 18–22 h at 22°C and 220 rpm, and then purified by Ni–NTA agarose (Novagen, Germany). The metallization of the proteins (only for holo-SOD1) was further performed by serial dialysis, consisting of three main phases of metal removal, metal charging, and unbound metal removal as previously described [16]. As the final point, protein concentration was set through the Bradford method using the BSA as the standard curve, and then the analysis was carried out on the 12.5% SDS-PAGE gel.

Enzyme activity assay

The SOD1 capacity to prevent pyrogallol autoxidation was utilized to measure its activity [40]. Overall, the reaction mixture, with the total volume of 300 μl and 0.2 μg/ml recombinant protein in 50 mM Tris–HCl buffer, 0.1 mM pyrogallol dissolved in 10 mM HCl, and 1 mM ethylenediaminetetraacetic acid (EDTA) at pH = 8.2, was measured at 420 nm for 5 min at 30-s intervals, using a microplate spectrophotometer (BioTek Epoch, the USA). The quantity to avoid pyrogallol autooxidation by 50% per min was further defined as the one unit of SOD activity, and then some individual activities were computed for each sample, as mentioned earlier [16].

Fluorescence spectroscopy

A fluorescence spectrometer (FP-8300, JASCO Inc., the UK) was used to conduct the fluorescence-related investigations. Following the dialysis of pure SOD1 in phosphate buffer (20 mM at pH = 7.4) with the protein concentration of 20 μg/mL, intrinsic fluorescence was measured within the final volume of 2 mL. Intrinsic fluorescence measurements were recorded with excitation fixed at 295 nm and emission spectra between 300 and 450 nm. The excitation and emission slit widths were set at 5 and 10 nm, respectively. Moreover, the ANS probe was utilized for extrinsic investigations to track conformational changes in proteins once bound to hydrophobic patches. The purified enzyme and ANS concentrations under amyloid-inducting conditions (50 mM Tris–HCl, 50 mM DDT, 0.2 M KSCN, and pH = 7.4 at 37 °C) were 10 and 30 μM, in that order. ANS emission was scanned a ranging from 370 to 700 nm involving some excitation wavelength equal to 350 nm. The excitation and emission slit widths were set at 5 and 10 nm, respectively. Entrance and exit slits were set at 2 nm.

Fourier transform infrared (FTIR) spectroscopy

Applying potassium bromide (KBr) tablets, the FTIR spectroscopy (BRUKER TENSOR 27, Germany) was fulfilled to analyze the protein secondary structure. The infrared spectrum was then set at the wave number of 400–4000 cm^˗1, with the resolution of 4 cm⁻¹. The apo/holo-WT-SOD1 and L126S protein samples at the concentration of 25 μM, under amyloid-inducing conditions (50 mM Tris, 50 mM DDT, 0.2 M KSCN, and pH = 7.4 at 190 rpm) for 72 h at 37 °C, were also incubated and 10 μl of each sample was ultimately used for the FTIR measurements. Subsequently, the buffer baseline was subtracted before taking each spectrum [41]. The peak assignments were accordingly made by the formerly defined spectral components associated with various secondary structure elements [42]. The raw spectra within the amide I region (1600–1700 cm^˗1) were further processed, using the OriginPro 2021 software package. During the FTIR curve fitting approach to analyze the amide I band components, the Fourier self-dissociation (FSD) and second derivative spectra were employed to visualize the overlapping peaks in the amide I region.

Apo/Holo-SOD1 chemical denaturation: thermodynamic (TD) stability

Here, the effect of GdnHCl concentration on the equilibrium unfolding transition of WT-SOD1 and L126S mutant in the apo/holo form at the protein concentration of 20–30 μM, 50 mM DTT (0–7 M), GdnHCl at 37 °C, 20 mM phosphate buffer, and pH = 7.4 was determined by a fluorescence spectrometer (FP-8300, JASCO Inc., the UK). Fluorescence measurements were recorded with excitation fixed at 295 nm and emission spectra between 300 and 450 nm. The excitation and emission slit widths were set at 5 and 10 nm, respectively. The Gibbs free energy of denaturation (ΔG⁰) was further established directly for both states of Native (N) ↔ Denatured (D) [43], as a transition, through plotting the data for fluorescence intensity vs. GdnHCl concentration. Of note, the data were expressed with respect to unfolded F_D fraction, using Eq. 2 below:

$${\text{F}}_{{\text{D}}} \, = \,{{\left( {{\text{Y}}\,{-}\,{\text{Y}}_{{\text{N}}} } \right)} \mathord{\left/ {\vphantom {{\left( {{\text{Y}}\,{-}\,{\text{Y}}_{{\text{N}}} } \right)} {\left( {{\text{Y}}_{{\text{D}}} \,{-}\,{\text{Y}}_{{\text{N}}} } \right)}}} \right. \kern-0pt} {\left( {{\text{Y}}_{{\text{D}}} \,{-}\,{\text{Y}}_{{\text{N}}} } \right)}} \,$$

(2)

In this equation, Y refers to observed variable parameter and Y_N/Y_D indicates the variable characteristic values of folded/unfolded conformations, respectively. The free energy difference between the folded as the fully native and unfolded states (viz., ΔG⁰) was then computed via Eq. 3 [44]:

$$\Delta G^{0} \, = \, - \,RT{\text{ln}}K\, = \, - \,RT{\text{ln}}\,\left[ {{{F_{D} } \mathord{\left/ {\vphantom {{F_{D} } {\left( {{1} - F_{D} } \right)}}} \right. \kern-0pt} {\left( {{1} - F_{D} } \right)}}} \right] \,$$

(3)

Here, K shows equilibrium constant, R stands for gas constant (8.314 J mol⁻¹ K⁻¹), and T represents absolute temperature. As well, ΔG⁰ varies in a linear manner with the GdnHCl denaturant concentration absorbance in the limited region [45].

$$\Delta G^{0} \, = \,\Delta G^{0}_{{{\text{H2O}}}} \,{-}m\, \times \,\left[ {{\text{denaturant}}} \right]$$

(4)

Of note, ΔG⁰_H2O stands for protein structure durability in a natural way with no denaturant, and m is determined by ΔG⁰ relevance to denaturant concentration [46].

hSOD1 aggregation characterization by ThT fluorescence

Utilizing the ThT fluorescence techniques, the kinetics of fibrillation was investigated based on the growth in fluorescence quantum efficiency of ThT after binding to amyloid fibers. The aggregated SOD1 protein samples were further incubated under the conditions inducing amyloid aggregation (namely, 50 mM DTT, 50 mM Tris–HCl, 0.2 M KSCN, pH = 7.4 at 37 °C, 190 rpm, at time interval of 0–72 h). The concentration of the apo/holo-SOD1 fibrils (30 μM dimer, 25 mM phosphate buffer at pH = 7.4) was then recorded. Afterward, ThT (20 μM) was added to the incubated samples at different intervals. Fluorescence was also as excitation and emission at 444 and 485 nm, respectively. The excitation and emission slit widths were further set at 5 and 10 nm, in that order. The data reported were thus the mean measurements within three replicates. As reported before [47], the kinetics of fibrillation was established using the formula below:

$${\text{F }} = \,\frac{Fmax}{{1\, + \,\exp \,\{ - \left[ {\left( {t - tm} \right)/\tau } \right]}}$$

(5)

In this regard, F and F_max refers to fluorescence intensity at time t and end time of ThT signal recording, respectively. As well, t shows characteristic time constant and t_m stands for time through which half of amyloid aggregates occur. F_max and τ are also floating parameters ion accordance with the best-fit method. Moreover, k_app denotes apparent rate constant, given by 1/τ for increase in length of fibrils. Thus, t_D = t_m—2τ is utilized to compute delay time (t_D).

Transmission electron microscopy (TEM)

The SOD1 fibrillation samples were obtained once fibrillation reaction reached saturation and used for TEM. At different time intervals, 5 μL of aggregated samples were accordingly extracted and adsorbed onto carbon-coated 300-mesh Cu grids for 5 min. Then, negative staining was performed by means of 2% uranyl acetate and air-dried after 20 s. The samples were further checked using a transmission electron microscope (Philips, EM 208S, the Netherland) at the accelerating voltage of 100 kV.

Results and discussion

Analysis of single nucleotide polymorphisms (SNPs)

SNPs constitute the most frequent sources of genetic variation in the human genome and they serve as a key indicator in genetic research. Of note, protein stability can be shaped by nonsynonymous DNA mutations that modify amino acid sequence and potentially prevent required structural modifications for proper protein function [48]. The interactions that occur between amino acids and their surroundings thus produce a stable three-dimensional (3D) structure of proteins. The protein-environment system in the minimum Gibbs free energy protein folding (ΔG) also consists of interaction energies within the protein, including hydrophobic interactions, hydrogen bonding, and electrostatics, etc. as well as entropic contribution, such as hydrophobic effect and protein configuration [49]. The difference in ΔG is accordingly applied to measure the effect of non-synonymous variants on protein stability. Considering ΔΔG^u as the measure of apparent free energy difference between mutant and WT-type proteins, ΔΔG^u = ΔG^u _mutant—ΔG^u _wild-type, meaning the difference in unfolding free energy between mutant and WT-proteins [50]. Consequently, stability changes were predicted and conformational instability associated with hSOD1 was estimated using different algorithms, viz., i-Stable, I-Mutant2, and DUET, explaining the difference in protein conformational stability induced by substitution mutation. The results of integrated servers computed in this line for L126S mutation as presented in Table 1 showed destabilizing effects on the WT-SOD1 structure, which help understand the way mutation influences hSOD1. The role of missense mutations in changing protein stability had been already established [49].

Table 1 The performance effect of mutation on protein stability in silico analysis of SOD1 mutant

The computational techniques of Predict-SNP were additionally practiced to assess the way amino acid alterations could affect the protein function. The outcomes of the servers and computational algorithms could be thus used to forecast if an option was harmful or neutral. A deeper understanding of the impact of non-synonymous mutations on protein stability was further required to connect their involvement in various disorders. The Predict-SNP analysis findings for the L126S variant accordingly indicated some harmful consequences and the projected confidence score (Table 2) was 0.87. To ensure exceptional dependability, some criteria were taken into account. The results were in agreement with previous reports for other mutations in protein stability [26].

Table 2 Predictions and confidence scores were obtained from the best six algorithms of the PREDICT-SNP server for the SOD1 mutants

MD-based simulation analysis

Notably, the changes that occur in protein amino acids could alter protein folding, stability, and biochemical functions and then bring negative outcomes. MD simulations were thus used to reflect on the effect of mutation on SOD1 protein and reveal its flexibility and structural deviation. Therefore, the simulations were performed on the apo/holo-WT-SOD1 and L126S mutant for 150 ns at 310 K to find the structural dynamic changes.

First, root-mean-square deviation (RMSD) was considered to determine the difference between the initial structural conformation and the final position of Cα atoms of the protein backbone. The changes arising during the protein simulation could be accordingly used to determine its structural conformational stability [51]. The RMSD results further indicated that the point mutation in the SOD1 form changed the overall basis of the protein. In spite of this, the RMSD of the holo-WT-SOD1 and mutants demonstrated less change during the simulation. As well, there was no significant difference for simulation between the holo-WT-SOD1 and the L126S mutant. The disruption seen in the key loops in the simulation also resulted in a fast rise with a steep slope for the apo-L126S as compared to the WT form, which suggested structural instability. The mean RMSD values for the holo-WT-SOD1 and L126S mutant were also equal to 0.23 and 0.23 nm, respectively (Fig. 1a). Besides, such values for the apo-WT-SOD1 and L126S mutant were 0.23 and 0.28 nm (Fig. 1b). In general, metal deficiency and mutations could reduce the structural stability and aggregation of proteins.

Fig. 1

a, b RMSD as a function of time. The RMSD for the backbone atoms of the apo/holo WT-SOD1 and mutant is shown as a function of time. c, d RMSF for each residue of apo/holo WT-SOD1 and mutant. Root-mean-square fluctuation for each residue of hSOD1. e, f Radius of gyration as a function of time. The radius of gyration of the apo/holo WT-SOD1 and the mutants during the MD trajectory is shown. g, h Hydrogen bonds (H-bond) function of time. The hydrogen bonds of the holo/apo-WT-SOD1 and mutant during the MD trajectory are shown. i, j The SASA of the holo/apo-WT-SOD1 and mutant during the MD trajectory are shown

Next, root-mean-square fluctuation (RMSF) was employed to quantify the residual flexibility of the apo/holo-WT-SOD1 and L126S mutant proteins. Of note, RMSF was among critical quantities to measure the deviation of a set of atoms in their average positions in a structural system through MD simulations [51]. The results accordingly revealed that the mutation altered the overall flexibility of the protein as compared to WT, and such fluctuations seemed to be greater in the apo-SOD than the holo one. Moreover, reduced flexibility was observed in the functionally important loops of 49–83 (that is, Zn loop/loop IV) and 121–142 (namely, loop VII/electrostatic loop) in the holo-L126S mutant as compared to the WT-SOD1. Furthermore, the rising trend in the fluctuations in the apo-L126S mutant indicated a disruption in these loops due to higher fluctuations. More fluctuations in the electrostatic loop might be thus attributed to metal deficiency and mutations that could significantly change flexibility [52]. Moreover, the fluctuations in the apo forms suggested that the loss of metal ions from their active sites was likely to boost up the driving force for aggregation propensity in the apo-SOD1 form as compared to the holo one. The results also established that the mutation effects could give rise to severe deformation or disruption of the given loops and make some changes in protein flexibility. This could then cause conformational stability loss in the protein. Besides, the removal of metal ions from the active sites could lead to further loosening [8]. Such a reduction in flexibility was more evident in the apo-SOD1 form. The high flexibility of the electrostatic loop could form extended or open conformations in the SOD1 variants. Furthermore, the mean RMSF value exhibited lower flexibility in the L126S mutant (0.066 nm) in comparison to the WT-SOD1 (0.075 nm) in the holo-SOD (Fig. 1c), whereas the mean RMSF value showed greater flexibility in the L126S mutant (0.124 nm) as compared to the apo/holo-WT-SOD1 (0.085 nm) (Fig. 1d). The simulation outputs accordingly revealed that the SOD1 protein could lose residual flexibility and its structural strength with no Zn ions, thereby highlighting the significance of the binding of such ions in maintaining the structural integrity of the protein [53]. The results thus revealed the different nature of the dynamic properties of the SOD1 variants. In this context, the native SOD1 underwent some changes in flexibility in metal binding and electrostatic loops as compared to the apo/holo mutant. Moreover, the electrostatic loop showed the greatest change in conformational fluctuations, as previously reported, viz., energetic electrostatic loops could increase interactions and formation of SOD1 amyloid fibrils [54].

Correspondingly, protein stability and total number of intermolecular contacts were associated with the relative compaction of the protein structure, as illustrated by the radius of gyration (Rg) [55]. Accordingly, a rise and a fall were seen in density fluctuations the holo and apo mutant proteins, respectively. During the simulations, the mean Rg values for WT and L126S (holo form) were 1.44 and 1.46 nm, respectively (Fig. 1e). In contrast, the protein compaction values for WT and mutant (apo form) were equal 1.44 and 1.41 nm, in that order (Fig. 1f). Low Rg values in the apo-SOD1 thus indicated higher protein density, attributable to metal deficiency and increased intramolecular interactions. The RMSF, Rg, and RMSD values of the apo-SOD1 further fluctuated more than those of the holo-SOD1, implying that the former was more prone to aggregation than the latter owing to the removal of metal ions from its active sites, as they were in charge of the exceptional stability of the protein while it was operating at the maximum capacity.

Besides, hydrogen bonds were one of the most common non-covalent interactions in proteins, playing a key role in protein structure and function [56]. They were also the first bonds to respond to the structural perturbations induced by different protein mutations [57]. The number of such bonds was thus calculated to find the possible effects of mutation on the conformational changes in the apo/holo-SOD1 forms. The mean number of H-bonds for the L126S mutant and holo-WT-SOD1 was equal to 106 and 100, respectively (Fig. 1g). However, a growing number of H-bonds were spotted for the apo-forms, the WT and mutant 103 and 109, respectively (Fig. 1h). Previous research had further revealed that pathogenic mutations in SOD1 were likely to alter such bonds in proteins and promote the formation of new bonds that help increase the probability of misfolding and aggregation [58]. In sum, higher intramolecular interactions, greater unfolding, and less stability of the apo-SOD1 structures as compared to the holo-SOD1 were reported. The intermolecular H-bonds formed in the apo-SOD1 in comparison to the holo-SOD were significantly expanded by mutation and metal deficiency, suggesting some changes in protein compaction. These results were consistent with the Rg parameter outcomes. Previous studies [26] had further shown that SOD1 mutants could augment the propensity of the protein to aggregate into oligomers and amyloid fibrils through the aberrant increase of H-bonds as well as hydrophobic interactions. In this context, the exposed surface area of protein structures accessible to solvent molecules, has been introduced as the solvent-accessible surface area (SASA), which is to assess the way proteins and solvents interact and determine the characteristics and activities of proteins [26]. The steady behavior of all simulations was thus expressed by the SASA values. Throughout the experiment, the variations in the apo-form had a decreasing trend, implying that the formation of a compact state of SOD1 was caused by the apo form. As depicted in Fig. 1i, the mean SASA values for the holo-WT-SOD1 and mutant were equal to 87.7 and 88.1 nm², respectively, whereas such values were 86.5 and 84.6 nm² for the apo-WT-SOD1 and mutant, in that order (Fig. 1j). Protein misfolding, linked to the ejection of hydrophobic side chains toward the protein core, also increased as the SASA value dropped, indicating a reduction in the exposed hydrophobic residue. These findings were also confirmed by the fluctuations of the Rg and H-bonds in the simulations, consistent with previous reports. The change in dynamics could thus destabilize the structure, leading to it misfolding and relative compaction along with increased dimer dissociation and aggregation. These results suggested that the apo-SOD1 was more likely to aggregate than holo one. The loss of metal ions also destabilized the dimer and produced an apo-monomer, thereby representing an aggregation entry point [14].

At the same time, the propensity of the secondary structure content per residue of the apo/holo-WT SOD1 and L126S mutants was computed using the Dictionary of Secondary Structure in Proteins (DSSP) algorithm (Table 3). As such proteins could play a key role in creating hazardous aggregates, secondary structural features like β-strands were necessary to understand the intrinsically disordered ones involved in NGDs. In this line, significant alterations were seen in the secondary structural characteristics of SOD1 upon the apo/holo mutation. The Zn-binding and electrostatic loop residues of the WT and mutant also showed the highest frequency of loop generation. Other forms also indicated a lower tendency for loop formation than the holo-WT. Furthermore, there was a notable rise in the β-sheet content when the apo-SOD1 was compared to the holo-SOD1. Hence, the secondary structure propensity demonstrated that mutation increased the β-sheet secondary structure content, as an important parameter in determining protein aggregation. The α-helix content further reduced in the apo/holo mutant. Of note, α-helix is a functionally important loop present in loops IV and VII of SOD1. The reduced α-helix content accordingly expanded loops IV and VII, giving rise to protein destabilization and more susceptibility to aggregation [59].

Table 3 Secondary structural analysis of apo/holo WT-SOD1 and L126S mutant

Apo/Holo SOD1 purification and enzymatic activity

To reach purity, the His-tagged SOD1 protein was initially loaded onto the Ni–NTA affinity column. Afterwards, its main characteristics were described on 12.5% SDS-PAGE to validate the molecular weight and expression the level of the protein. Using the SDS-PAGE gel, the purified SOD1 represented a single band with the molecular weight of about 16 kDa, which corresponded to the high-purity SOD1 protein produced (data not shown). After the protein expression and purification were completed, the action of the Cu chaperone for SOD (CCS), as a protein found in eukaryotes, was considered vital for proper SOD1 metallation [60]. However, the metals were detected in vitro after protein purification by prolonged incubation at low temperatures, allowing the SOD1 expression in prokaryotic cells as well as SOD1 metal manipulation [61]. Even though the protein could be expressed in bacteria, no appropriately acetylated N-terminus as a post-translational modification at the SOD1 dimer interface was required. Upon confirming the expression, purification, and metalation of native and mutant proteins, the Flame atomic absorption spectroscopy (FAAS) as an analytical method was used for the holo-variants containing all metals and then on the apo-variants after dialysis. This technique could provide accurate results when using pure protein in solution [62], (data not shown).

Although SOD1 had a high degree of stability, point mutations could lead to structural instability, thereby promoting oligomerization and aggregation. Misfolding and aggregation of proteins were also associated with their loss of stability. Certain mutations in the metal-binding loop of SOD1 linked to ALS could thus impact the residues coordinating Cu or Zn and minimize catalytic activity. The holo-WT-SOD1 and L126S mutants accordingly had enzyme activities of 5858 ± 128 and 2850 ± 218 U/mg, respectively. On the other hand, such values were 1950 ± 277 and 1133 ± 236 U/mg for the apo-WT-SOD1 and L126S mutation, in that order. (Standard deviations were of the experimental values of three independent experiments). The results further revealed lower enzyme activity for the apo-SOD1 as compared to the holo-WT-hSOD1. Even if some H bonds and intramolecular interactions were of importance for SOD1 folding and stability of, mutation and metal deficiency could aberrantly increase hydrophobic and H bonds and then decrease enzymatic activity in SOD1 [63]. During the simulations, loops IV and VII also showed larger variations in the apo-SOD1 as compared to the holo-SOD1 one, suggesting a significant malfunction in these loops based on the RMSF data. Moreover, the flexibility changes were in line with loop activity loss, which eventually resulted in the complete loss of SOD1 catalytic function. As documented, loop VII could produce an ideal electrostatic field to absorb superoxide anion radicals. The conformational shift induced by this mutation could then alter the orientation of the metal ligands in the mutation reaction, and consequently reduce the catalytic activity of the reaction [64]. As a whole, these data implied that metal-binding as key factor in SOD1 aggregation was essential for the enzymatic activity of this antioxidant. In detail, demetallation and aberrant metal binding could promote SOD1misfolding and aggregation, suggesting the possible role of metal binding in pathological SOD1 aggregation [64]. Considering the partial or localized feature so the changes, they might have a significant effect on enzyme structure and activity.

Fluorescence spectroscopy analysis

As previously confirmed, the intrinsic fluorescence emission of tryptophan (Trp) could be measured using fluorescence spectroscopy. The changes in the environment of Trp were thus capable of altering fluorescence intensity and peak [65]. The intrinsic fluorescence of the apo/holo WT and mutant was investigated under pre-induction conditions (protein with a concentration of 0.02 mg/ml in 20 mM phosphate buffer pH 7.4 at 25 °C). The increase was thus observed in the intrinsic fluorescence of the holo-L126S mutant as compared to holo-SOD1, indicating that only Trp 32 was placed in the hydrophobic environment (Fig. 2a). On the other hand, the drop in the intrinsic fluorescence emission of the apo-L126S as compared to the apo-WT showed that the polarity of the environment around the Trp residue increased (Fig. 2b). In other words, it was placed in a more hydrophilic environment and was more exposed to the solvent.

Fig. 2

Structural characteristics of SOD1 by intrinsic and ANS fluorescence. a, b Intrinsic fluorescence emission of holo/apo WT-SOD1 and L126S mutants in phosphate buffer (0.02 mM, pH 7.4). c, d Intrinsic fluorescence emission of holo/apo WT-SOD1 and L126S mutants under amyloid-inducing conditions (50 mM DDT, 50 mM Tris–HCl, 0.2 M KSCN. pH 7.4) and final protein concentration (0.02 mg/ml at 37 °C). All data was obtained from replicates of two purified protein samples, each analyzed in two independent replicates. Control = the apo/holo WT and mutant was investigated under pre-induction conditions (protein with a concentration of 0.02 mg/ml in 20 mM phosphate buffer pH 7.4 at 25 °C). e, f The ANS fluorescence emission of holo/apo-WT-hSOD1 and L126S mutants in phosphate buffer (20 mM), under amyloid-inducing conditions (50 mM DDT, 50 mM Tris–HCl, 0.2 M KSCN. pH 7.4) and final protein concentration (0.02 mg/ml at 37 °C). The concentration of the ANS in the enzyme solutions was 30 µM, and the molar ratio of protein to ANS was 1:30. The standard deviation was from the experimental values of two independent experiments

To realize the effect of mutation on tendency to form amyloid aggregates, the intrinsic fluorescence of the apo/holo WT and mutant was investigated under reducing conditions (viz., 50 mM DDT, 0.2 M KSCN, 50 mM Tris–HCl, 0.02 mg/ml protein concentration, and pH = 7.4 at 37 °C). Considering the holo-L126S, the decline in fluorescence emission indicated a growth in the polarity of the Trp residue, i.e., the Trp residue was exposed to the solvent and oriented toward a more hydrophilic environment (Fig. 2c). The intrinsic fluorescence results were accordingly in agreement with those of the MD analysis. On the other hand, the rising trend in fluorescence intensity in the apo-L126S mutant, as compared to the WT, was ascribed to the coverage of the available hydrophobic surfaces, inducing the binding of the monomers that is the simultaneous placement of β-sheets as well as amyloid aggregation due to the high-concentration of DTT (Fig. 2d).

To examine whether the mutations caused some changes in SOD1 hydrophobicity and then monitor the exposure of hydrophobic pockets in β-sheet structures during aggregate formation, ANS was applied as a charged hydrophobic fluorescent molecule, which could help to detect compacted, partially folded intermediates of the protein population [66]. The ANS binding to the intermediate levels of the hSOD1 mutants could be thus the result of a looser conformational state, as confirmed by the crystal structure of the hSOD1 mutant [67]. The hydrophobicity of the protein surface could further contribute to some steps of the SOD1 aggregation, and change based on the SOD1 mutation as well as metallization and demetallization [68]. For this reason, the effect of mutation and metal deficiency on the propensity of SOD1 to form protein aggregates under amyloid-inducing conditions (50 mM DDT, 200 mM KSCN, 50 mM Tris–HCl, at pH = 7.4) and final protein concentration (20 mg/ml at 37 °C) were evaluated by the ANS fluorescence. ANS without protein was used as control. All apo-SOD1 forms, in contrast to the holo-SOD1, revealed spontaneous aggregation followed by conformational changes in the course of the experiment. According to the preliminary evaluations (Fig. 2e, f), there was a significant increase in the ANS fluorescence intensity for the apo-SOD1 form as compared to the holo-SOD1, indicating a higher level of hydrophobicity in the protein. In these conditions, the S–S-reduced apo-WT-SOD1 and mutant were expected to monomerize and have disorders in loops IV and VII [69]. The increased ANS fluorescence for the mutants in this way suggested that the impacts of these substitutions on the backbone structure or other localized consequences could be differentiated by the ANS probe under the solution conditions. The hydrophobic patches and/or contacts were not fully incorporated into the holo-SOD to support various steps of protein aggregation because the holo-form with lower ANS fluorescence had less potential for aggregation than the apo-SOD. In view of that, the partial or total unfolding or even structural disturbances were prerequisites for protein aggregation. The ANS fluorescence further augmented and subsequently saturated for all apo-SOD1 variants (namely, the WT and mutant), implying that structural rearrangements and aggregation occurred in all apo-SOD1 variants, but not the holo-SOD1. However, the conformational alterations in the intermediates and products during the misfolding cascade in ALS-causing SOD1 mutations suggested that the concealed epitopes in the apo-SOD could be exposed more quickly than the holo-SOD. In addition to Cu/Zn removal from the SOD1 of interest, as a dimeric protein wherein each monomer was associated with hydrophobic interactions, the reduction of intramolecular S–S bonds or aggregation in the physiological conditions, mainly for the protein mutant was enough. Considering the ANS data, the DTT-mediated decrease in intramolecular S–S bonds induced gross conformational changes, followed by protein assembly mostly through intermolecular hydrophobic interactions. Henceforth, the findings supported the idea that misfolding due to metal deficiency might facilitate aberrant interactions or hydrophobicity of SOD1 with itself or other cellular components and accordingly contribute to neurotoxicity, as reported earlier [63].

FTIR spectroscopy

FTIR, as one of the techniques to deal with the secondary structure of proteins, was used to examine misfolding and aggregate formation [70]. Accordingly, the formation of β-sheet structures in the apo/holo WT-SOD1 and mutant under destabilizing conditions was confirmed after 72 h of incubation (Fig. 3). To distinguish types of β structures and determine amyloidogenic conformers, the amide I bands, the FTIR spectra were also widely used. Proteins could have a prominent vibrational band in the region of 1700–1600 cm⁻¹, coming from the stretching vibrations of the C = O peptide bond, with the highest vibrational absorption. Amyloid-like aggregates were further formed by intermolecular β-sheets due to protein aggregation, whose specific IR signature included an increase in 1618 ± 10 cm⁻¹ region [71, 72]. At the same time, the native β-sheets could produce an absorption peak in the 1630–1641 cm⁻¹ range. The amide I band is located between 1600 and 1700 cm⁻¹ and is closely connected to the backbone conformation peak 1611–1630 cm⁻¹ is assigned as cross β-sheet [73]. Investigating the amide I bands; a parallel arrangement was distinguished from the anti-parallel arrangement of the β-strands in protein aggregates. The parallel β-sheets also displayed an elevated component at 1630–1641 cm⁻¹ [74]. According to Table 4, there was a peak in the 1625 cm⁻¹ range for the holo-WT-SOD1, specific to the parallel structures of the β-sheets. The results of replication and control are shown in Table 1S and 2S. The changes in the β-sheet content were further observed in the holo-L126S mutant. Moreover, a mixture of parallel and intermolecular β-sheet structures was distinguished for apo-WT-SOD1 and apo-L126S, indicating the dominance of the β-sheet structure during aggregation. Besides, the FTIR spectroscopy outputs established that SOD1 with and without metal entailed different changes in the natural β-sheet structure of the protein owing to the aggregation method. Accordingly, the changes in the secondary structure in the apo/holo forms and the formation of amyloid aggregates indicated the structural rearrangement of SOD1, which were consistent with the β-sheet structure nature in the SOD1 aggregates, as suggested by the secondary structures based on the MD simulations. The results of replication and control are shown in Fig. 1 and 2S.

Fig. 3

Results of a Fourier self-deconvolution (FSD) analysis on the amide I (1700–1600 cm⁻¹ region) for a holo WT-SOD, b holo L126S mutant, c apo WT-SOD and d apo L126S mutant. The assignment of different peaks to different secondary structure elements is shown. Protein concentration was 25 μM under destabilizing conditions in 20 mM phosphate buffer (pH 7.4 at 37 °C and after 72 h of incubation). All data were obtained from replicates of two purified protein samples, each of which was analyzed in two independent replicates

Table 4 Assignment of amide I band components and percentages of secondary structure contents in apo/holo WT-SOD1 and mutant

Comparative TD stability analysis

A metal deficiency, reduced S–S bonds, and some SOD1 mutations could undermine protein conformation, i.e., lead to SOD1 destabilization, and then increase misfolding and aggregation propensity [23], the unfolding of the native chemical could frequently occur first during aggregation. Furthermore, the main cause of lower TD stability in some proteins, which permits unfolding and aggregation to happen following a comparatively little alteration under environmental conditions, could result in protein instability to form aggregates or amorphous fibrils. For structurally distinct proteins with varying α-/β-structure contents, a completely unfolded conformation produces aggregates and fibrils [75]. Given the extremely stable protein structure of SOD1 and its high potential for aggregation under inducing conditions associated with the L126S mutation, it is sensible to assume that metal loss or mutation under reducing conditions induces conformational changes and intramolecular S–S bonds significantly contribute to these changes. This implies the changes in stability and flexibility of decreased and/or mutant SOD1 are the possible sources of its altered aggregation behaviour. The most common technique to reach TD stability in proteins could be thus denaturant-induced unfolding. In this line, GdnHCl was employed as an effective chemical denaturant for practicing chemical denaturation on the holo/apo-SOD1 form and then investigate the impact of reduced S–S mutation on SOD stability. For the reduced proteins (apo/holoSH), a profile of the intrinsic fluorescence emission variations vs. GdnHCl concentration was acquired for the apo/holo-SOD1. As a measure of a globular module (like, protein, domain, and so forth), ∆G⁰ was calculated with the assumption that the native module (N) might directly and reversibly convert into the denatured state (D). As mentioned earlier, the given process could be examined by devoting more attention to the changes in absorption and emission and then computing ∆G⁰, with respect to the two-state mechanism for the transitions [76]. Ahead of normalization, the data were further analyzed based on the two-state model. As illustrated in Fig. 4a, b and Table 5, there was a simultaneous increase in intrinsic emission intensity in all SOD forms upon a growth in GdnHCl concentration. Moreover, their curves display a common transition from the native to the denatured state. At the denaturant concentration, ΔG⁰ was also computed. The ΔG⁰ dependence on GdnHCl concentration was of the linear type (Fig. 4c, d). The easiest way to estimate conformational stability when there was no denaturant, ΔG⁰_(H2O), was that the linear dependence could continue up to the concentration of zero. Figure 4 displays the fraction of the fully unfolded protein as the function of denaturant concentration for the reduced (apo/holoSH) proteins. Further analyses yield the estimation of two interrelated thermodynamic parameters: the m value (the dependence of protein stability on the denaturant concentration) and ΔG⁰ _(H2O) (the unfolding Gibbs’ energy extrapolated to zero concentration of denaturant). These parameters are presented in Table 5. Besides, the apparent ΔG⁰_(H2O) values of the reduced proteins (apo/holoSH) were equal to 13 ± 0.33 (holo-WT), 7 ± 0.65 (holo-L126S), 9.4 ± 0.25 (apo-WT), and 5.2 ± 0.36 (apo-L126S) kJ mol⁻¹, in that order. Accordingly, the conformational stability of the apo forms significantly reduced as compared to the holo-SOD1. Furthermore, the reduced form of the apo/holo-SOD1 had a smaller ΔG⁰ value in comparison to the holo-WT protein. In spite of the presence of the apo-SOD1 species in both WT and mutant forms with lower conformational stability as compared to holo-WT-SOD1, there was a greater potential for aggregation, suggesting that induction was sufficient to initiate SOD aggregation. Of note, the SOD conformational fluctuations could be strongly dependent on metal ion binding, so metal deficiency could accelerate dimer dissociation and/or monomer unfolding [77]. Conversely, the lowered SOD1 variants showed a drop in the ΔG⁰ levels. To start the SOD1 misfolding and aggregation, demetallation and S–S reduction needed to occur at the same time because none could considerably lower protein stability on their own. Given that toxicity could be correlated with the quantity of aggregated species, this implied that there were more unstable forms of SOD to aggregate and cause diseases; e.g., due to metal loss, monomerization, or reduction of S–S bonds [78].

Fig. 4

Respectively, fluorescence emission spectra and thermodynamic stability of a–b apo/holo WT-SOD and L126S mutant in the range of 0–7 M GdnHCl and 20–30 µM protein concentration in 20 mM phosphate buffer under amyloid induction conditions (DDT 50 mM, Tris–HCl 50 mM, KSCN 0.2 M, pH 7.4, at 37 °C with and without agitation at 190 rpm. c and d show free energy changes for chemical denaturation of DTT-mediated reduced apo/holo WT-SOD1 and variant. The standard deviation was from the experimental values of two independent experiments

Table 5 Calculated thermodynamic stability parameters for the apo/holo SOD1 variants protein, against GdnHCl, in the presence of DTT

ThT fluorescence assay

The ThT fluorescence intensity end-point level could provide more information about mature fibrils. In view of that, ThT could attaches in a parallel manner to the long axis of the fibril and intercalate to the repeated side-chain contacts running across the β-strands inside the β-sheet layer, with regard to the growing number of structural models for amyloid fibrils [79]. The apo/holo-WT-SOD1 and its mutations were accordingly subjected to incubation for 0–144 h at 37°C, with and without agitation (190 rpm) at 50 mM DTT, 0.2 M KSCN, and 50 mM Tris at pH = 7.4. When KSCN was swapped out for NaCl, no amyloid developed, indicating that its effects extended beyond its ionic strength due to chaotropic characteristics. Subsequently, the amyloid structure had amyloid-like aggregates, as confirmed by ThT fluorescence and TEM. Figure 5 shows the representative fluorescence time courses for both the mutant and apo/holo-WT-SOD1 forms, and Table 6 outlines the kinetic data. The findings revealed that amyloid aggregates could be formed by WT-SOD1 and mutant in the presence of high DTT concentrations and metal deficiency. There was also a lag period before the rise in ThT fluorescence in the SOD1 fibrillation. Moreover, the lag phase duration varied between the apo/holo WT-SOD1 and mutant forms. Comparisons disclosed that the apo-SOD tended to shorten lag phases and accelerate fibrillation more than the holo-SOD. ThT fluorescence further increased after this phase with varying degrees for the apo/holo WT and mutant, explained by (i) variations in the quantity of attached ThT molecules and (ii) variations in the quantum yield or fluorescence intensity of fibril-bound ThT [80]. Discrepancies in the dynamics, structures, and morphologies of fibrils could accordingly cause changes in binding affinity and quantum yield. To give an example, the interactions between fibrils, e.g., fibril matting [81], could diminish ThT binding via lower solvent accessibility or secondary structure disruptions. This change in lag phase time was thus consistent with other reports fulfilled under different conditions [25, 82]. Nonetheless, the mutations inducing fALS might fail to invariably decrease latency and/or an increased tendency for SOD1 aggregation. The kinetics of aggregation n in amyloid fibrils in this way exhibited a lag phase, succeeded by the initial rise in ThT fluorescence intensity, which was more evident in the apo-SOD than the holo-SOD, as well as a subsequent fall in ThT intensity. This might be attributed to certain modified or inaccessible ThT binding sites resulting from mat-like fibril aggregates. The loss of metal ions might be the underlying cause of this variation in aggregation kinetics in the apo-SOD form. Ultimately, the plateau phase started when the ThT fluorescence intensity reached its maximum. This stage was short-lived because amyloid fibrils could often form web-like network structures. The study results thus endorsed the theory that SOD1 aggregation could happen via a process with dimer separation, apo-monomer aggregation, metal ion loss from monomers, and weakened dimer interface [83, 84]. According to previous studies, multiple aggregation pathways existed, viz., those leading to amorphous aggregates and some forming amyloid aggregates. Furthermore, the results showed that hSOD1 formed amyloid aggregates under different conditions, including agitation, temperature, ionic strength, and physiological pH. Besides, there was the potential of WT and mutant in the presence and absence of metal to form amyloid structures under amyloidogenic conditions. These findings showed that hydrophobic contacts could play a major role in the kinetic shifts between protein aggregates. Upon promoting primary nucleation events by increasing the local concentration of SOD1 at the surface and/or changing the SOD1 shape or its oligomers to accelerate nucleation, these hydrophobic contacts were likely to speed up fibrillization. These were in line with earlier research on the significance of SOD1 oligomerization or fibrillation in ALS pathogenesis [16, 23, 25, 82]. As shown in Fig. 6, the presence of hSOD1 amyloid-like aggregates was validated by TEM.

Fig. 5

Kinetics of a holo WT-SOD1 and L126S mutant b apo-WT-SOD1 and L126S mutant amyloid aggregation Formation using ThT fluorescence incubated in the presence of 30 μM protein in 20 mM phosphate buffer, ThT concentration 20 mM and under induction conditions (DDT 50 mM, Tris 50 mM, KSCN 0.2 M) pH 7.4 at 37°C with and without agitation at 190 rpm. The standard deviation was from the experimental values of two independent experiments.

Table 6 Kinetic parameters of amyloid aggregation formation from apo/holo SOD1 and mutants in 50 mM Tris, 0.2 M KSCN, and 50 mM DTT at pH 7.4 with agitation at 190 rpm at 37 °C

Fig. 6

Electron microscopy images (TEM) of apo/holo SOD1(WT and L126S mutant) incubated in Tris 50 mM, KSCN 0.2 M, and DTT 50 mM at 37 °C, pH 7.4, 190 rpm. a holo-WT-SOD1 before incubation b 96 h after incubation c holo-L126S before incubation d holo-L126S 72 h after incubation e apo-WT-SOD1 before incubation f 72 h after incubation f₁ A sample with 100 nm magnification from the apo-WT after incubation. g apo-L126S before incubation h 72 h after incubation. I A sample with 150 nm magnification from the apo-L126S mutant after incubation. The black arrows point at oligomers (O), fibril (F) and protofibrils (P) (Scale bar = 100,150, 300, and 500 nm)

Amyloid aggregation TEM

Notably, misfolded proteins could form aggregates, including amyloid fibrils and amorphous aggregates. In this context, amorphous aggregates that have been similar to amyloid fibrils have been implicated in the development of some serious NGDs, wherein intermolecular interactions form amorphous aggregates and amyloid fibrils [3]. Over recent years, the propensity to aggregate has been related to the TD stability and structure of the protein, which is closely linked to aggregation kinetics. Although amorphous aggregates and amyloid fibrils have distinct morphology and thermodynamic stability [85] and are associated with aggregation kinetics, they are typically difficult to tell apart [86]. Therefore, it is necessary to investigate them simultaneously in computational and experimental studies. Unlike amyloid fibrillation, there is not sufficient information on the kinetics of amorphous aggregation. However, amorphous aggregation is likely to occur when proteins are denatured, as unfolded conformational changes with exposed hydrophobic surfaces increase the tendency to aggregate. Finally, the formation of amyloid aggregates acquired by incubation proteins under destabilizing conditions, using the ThT fluorescence assay, was confirmed by the TEM images, which predominantly showed different aggregate morphologies, such as protofibrils, oligomers, fibrillars, and amorphous aggregates. Of note, the evidence of the aggregation mechanism could be provided from the TEM images at different times during aggregation, but no morphological changes related to the formation of intermediate amyloid fibrils were observed in the holo-WT-SOD1 form (Fig. 6).

Accordingly, TEM analysis was performed for the WT-SOD1 at time zero (Fig. 6a). In this way, oligomeric and protofibril intermediates formed only after incubation for 96 h at 37°C in the holo-WT-SOD (Fig. 6b). In vitro studies had further established that metallated SOD1 had not produced amyloid-like aggregates at the neutral pH [87]. Figure 6c depicts the TEM morphology of the holo-L126S mutant at time zero. The samples were thus imaged after 72-h incubation at 37°C at the end of the lag phase, which paralleled the formation of intermediate amyloid fibrils and the maximum increase in fluorescence, and then accompanied by fibrils (Fig. 6d) or web-like networks. No significant aggregation was further seen in the apo-WT without incubation, suggesting the presence of protofibril forms and fibrils under inducing conditions (Fig. 6e). For example, apo-WT forms thinner fibrils (∼1–2 nm in width). Large amyloid aggregates were then formed following prolonged incubation (apo-WT forms relatively thicker fibers (∼3 to 5 nm in width) after incubation) [88]. After 72 h (Fig. 6f), as the time point in the elongation phase of most reactions, the morphology of the apo-WT transformed to amyloid aggregation in the same way as rod-like structures and fibrillar aggregates [89]. Moreover, there was a significant increase in the formation of fibrils. In the case of the apo-L126S mutant, with no incubation under destabilizing conditions for 72 h, amorphous grains and networks were spotted (Fig. 6g). After 72-h incubation, the amyloid network was detected along with a large network of amorphous filaments, fractal-like aggregates, and occasional fibrils, when most reactions reached the final plateau (Fig. 6h) [90, 91]. To encourage more aggregation, the metals from SOD1 could be removed while their intramolecular S–S link was maintained or their intramolecular S–S bond was reduced [8]. The apo-state was thus susceptible to self-association because Zn and electrostatic loops were very flexible and disordered. Apart from the S–S redox state, the demotivated SOD1 generated amyloid-like aggregates under physiological pH, ionic strength, and temperature. This implied that amyloid formation was sufficiently triggered by the removal of Cu and Zn [23]. The apo-SOD1 unfolding was also suggested by the examination of its structure and mobility as well as S–S-oxidized holo-SOD1. According to X-ray crystallography and hydrogen–deuterium exchange (HDX) investigations, the electrostatic loop could cause major abnormalities in the apo-SOD1. Furthermore, S–S-reduced apo-protein exhibited a greater degree of surface hydrophobicity and reduced mobility in vitro as compared to S–S-oxidized holo-protein. Recent MD-based simulations had further bolstered the theory that the building block of SOD1 aggregation is the local unfolding of metal-free and S–S-reduced SOD1 [92, 93]. Under various incubation settings, prominent morphological changes, such as, prefibrillar species, mesh-like networks, and long and short fibrils with or without branching fibrils were observed. Considering a wide variety of destabilizing circumstances, including metal shortage, S–S bond reduction, temperature rise, or DTT addition, the production of amyloid-like aggregates was a typical characteristic. The available data accordingly demonstrate that SOD has a considerable propensity to aggregation under all instability circumstances. Although it had a high ThT signal, all morphologies were practically amorphous instead of fibrillar, indicating that SOD1 structure was distorted into β-sheets in such circumstances and their structure was not the same as that of amyloids. These results were consistent with the morphology reported in previous research on SOD1 aggregates induced by unstable conditions [16, 19, 20, 23, 25]. Realizing the processes developing a disease and the way it progresses can facilitate genetic counseling, creation of novel therapies, and planning for more successful clinical trials. With regard to the ongoing clinical trials on the effectiveness of antisense SOD1 oligonucleotide-based treatments, this is relevant for SOD1 [94]. It is thus better classifying trial participants with fast and slow progression and then provide more accurate estimates of their expected survival rate based on the significant differences in survival times between patients with SOD1 variants and the linked structural characteristics discovered in this domain. Since long-term trials are needed to verify putative positive effects on slow-progressing variations, the accurate prediction of predicted survival is vital for trial design and interpretation of results. Overall, the study findings supported the idea that misfolding associated with metal deficiency could facilitate aberrant hydrophobic interactions of SOD1 with itself or with other cellular components, and then contribute to neurotoxicity [63]. With reference to the controversial observations using bioinformatics and in vitro experiments about the aggregation mechanisms and toxic effects of SOD1 variants associated with ALS, the roles of mutation and metal deficiency plus other physicochemical factors needed to be investigated to better understand it in vitro and in vivo. The present study was as a whole useful for researchers in line with a structure-based drug design effort to develop anti-aggregation compounds to treat the incurable form of ALS affecting human.

Conclusion

Some unfolding variables, such as mutation, metal deficiency, intra- and inter-S–S bond reduction, hydrophobicity, altered solution composition, β-sheet tendency, agitation, temperature, and pH can be the underlying causes of protein aggregation. Thus, understanding the mechanisms by which SOD1 misfolds and aggregates in neurons seems to be crucial for developing treatments that help halt ALS pathogenesis. The main structural components of SOD1 accordingly are binding to Cu and Zn, so its disruption may contribute to the ALS pathophysiology. Identifying variations (the apo/holo forms) is in this line a possible component in the study design and interpretation. Of note, this study was designed based the tendency of SOD1 variants to dissociate or aggregate under different incubation conditions and combinations. The data thus give prominence to metal ions in the misfolding and aggregation behavior of the SOD1 variants. The MD-based data analysis further demonstrated that the changes in flexibility, protein hydrophobicity, stability, and intramolecular interactions of the apo-SOD1 form were more evident than the holo-SOD. On the other hand, the apo-SOD1 entailed reduced enzyme activity and lower TD and/or kinetic stability as compared to the holo forms, attributable to metal deficiency in their structure. Besides, the DSSP and FTIR outputs confirmed higher tendency to form β-sheets in the apo-SOD form. Furthermore, the TEM images helped establish the intermediate properties of the amyloid and amorphous aggregates by the ThT under destabilizing conditions. The holo/apo aggregation in numerous characteristics of the SOD1 variants was also similar to those of other disease-associated proteins that could aggregate. They could form fibrillar and amorphous structures, exhibit a delayed phase caused by unfavorable nucleation and overcome by seeding, and grow rapidly through secondary nucleation. To sum up, the study findings provided universal insights into protein aggregation in illnesses, as a foundation for future research aimed at defining aggregation mechanisms, identifying inappropriate interactions, and creating vital therapeutic approaches to prevent toxic protein aggregation, including SOD1.