It has long been recognized that DA and its oxidation products can inhibit and even reverse AS fibrillation in vitro and in vivo, deviating the aggregation pathway towards spherical, soluble oligomers [96]. This effect does not depend on protein covalent modification and is, rather, associated to protein conformational changes [96, 97]. DA oxidation, too, is not absolutely required, as shown by control experiments, in which the effect is retained while inhibiting oxidation by NaBH4. DA-induced AS oligomers slow down fibrillation in a dose-dependent manner [98]. Circular dichroism (CD) and infrared (IR) spectroscopy indicate that the conformational ensemble of oligomeric AS is shifted in the presence of DA towards a predominantly disordered state with some elements of $\alpha$- and $\beta$-secondary structure, resulting in an average $\beta$-sheet content in between that of monomeric and fibrillar structures [96, 97]. The role of pH on the effects of DA on AS aggregation has also been investigated [99]. At acidic pH, a condition in which DA is stable, AS undergoes amyloid aggregation regardless of the presence of DA. DA impacts, instead, on AS aggregation at pH $>$7, an effect that is counteracted by antioxidants (ascorbic acid, glutathione, or homocysteine) and reducing agents ($\beta$-mercaptoethanol and dithiothreitol) to different extents, depending on the specific agent [99]. These results support the hypothesis that DA oxidation products are more effective than DA itself in inhibiting AS amyloid aggregation.

In-vitro characterization of serial AS deletion products and competition experiments by designed peptide have led to first identification of the main AS-DA interaction site within the C-terminal pentapeptide ${}_{125}$YEMPS${}_{129}$ [97]. That this sequence is required in vivo for AD-dependent inhibition of AS fibrillation has been demonstrated in SH-SY5Y cells expressing a multiple mutant carrying five amino acid substitutions in the 125–129 region, while controlling catechol levels by the expression of the biosynthetic, rate-limiting enzyme TH [100]. Such an evidence has been later extended to mice and Caenorhabditis elegans [26].

MD simulations of AS in aqueous solutions and clustering analysis on experimental NMR structures indicate that DA binds preferentially to the ${}_{125}$YEMPS${}_{129}$ region and to a few residues of the NAC region, establishing hydrophobic contacts as well as H-bonds [101, 102]. Those results also show that the AS-DA complex is stabilized by electrostatic interactions with the E83 residue of the NAC region. Mutation of this residue to alanine impairs fibrillation inhibition by DA, confirming an important role of this interaction in AS affinity for this ligand and hinting for the first time to a non-linear binding site [101]. The AS-DA complex in the different structural clusters appears to fluctuate around rather large average values (16.3–25.8 Å), indicating that the complex maintains large conformational flexibility, in spite of some induced structural compaction and induced structural order (Fig. 5, Ref. [102]). MD simulations on preformed AS44–96 protofibrils and DA have shown disruption of the Greek-key-like core by hydrophobic and electrostatic interactions at multiple AS-DA interaction sites [103].

In-vivo AS conformational properties and their response to DA concentration have been explored by the FRET-based assay for molecular proximity called fluorescence lifetime imaging microscopy (FLIM) [104]. This technique is based on the observation that the fluorescence lifetime of a donor fluorophore decreases with the 6th power of the distance from a FRET acceptor. Mouse and rat primary neuronal cultures were transfected by a doubly-tagged Myc-aSyn-V5 to probe the intramolecular distance between N- and C-termini by FLIM, based on immunocytochemistry with fluorescent labeled secondary antibodies. A first interesting observation was an inhomogeneous intracellular distribution of AS conformers, with more extended conformations in the cell body and more compact ones in the neurites. Furthermore, DA uptake significantly reduces the average distance, indicating that DA itself or some DA metabolite modulates AS conformational ensemble in vivo in favor of partially structured states. No effect was observed with DA agonists or antagonists. Very minor changes in CD spectra have been interpreted as an increase in random-coil and loss of $\beta$-structures, but these effects seem hardly significant [104]. Other reports by CD and/or FTIR failed to detect a significant effect of DA on monomeric AS secondary structure in solution [96, 105, 106].

DA-induced changes in AS conformational ensembles in vivo are mirrored by peculiar structural and functional features of the resulting oligomers [26]. Indeed, the oligomeric aggregates that accumulate in mice expressing the human A53T mutant, upon induction of catecholamine production, are larger (up to 122 Å) than in the control system (up to 65 Å) and do not display seeding activity on AS aggregation, as assessed by in-vitro Thioflavin T (ThT) assays and analysis of trans-synaptic spreading to motor cortex. Furthermore, AS oligomers formed in the presence of DA are known to be sodium dodecyl-sulfate (SDS)-resistant [100, 107] and this feature has been reported to be modulated by the concomitant presence of copper and DA [108]. DA-stabilized oligomers are suspected to have toxic effects by multiple possible mechanisms [109], such as inhibition of SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) complex [110], impairing chaperone-mediated autophagy [111], and causing mitochondrial dysfunction [112]. However, a recent article on SH-SY5Y cells and rat synaptosomes reports negligible toxicity of purified AS oligomers formed in the presence of DA [98].

In line with previous evidence, it has been shown that DA counteracts the effect of psychosine on AS conformation and aggregation in vitro [113]. Psychosine promotes AS fibrillation in ThT assays and has a fingerprint in ${}^{15}$N AS Heteronuclear Single Quantum Coherence (HSQC) NMR spectra, consistent with interactions with the C-terminal domain of the protein. AS incubation with both ligands abolishes these effects, although no comparison with the effect of DA alone is reported. The authors of that study speculate about interconversion between an “open” and a “closed” AS conformations, where the former (fibrillation-competent) would be stabilized by psychosine and the latter by DA.

Further NMR experiments have recently pointed out that ${}^{15}$N AS HSQC spectra are very sensitive to pH changes and that previous AS-ligand binding studies might be exposed to the risk of false-positive results, due to small pH changes induced by the components of commercial ligand formulations, rather than by ligand binding [114]. In particular, DA-induced changes in ${}^{15}$N AS HSQC spectra become evident only at pH 5 and not at pH 7, upon subtraction of the pH effect, under the experimental conditions employed in that study.

Native MS can provide important additional information in the attempt to describe disordered conformational ensembles [115, 116]. This technique captures conformational components coexisting in the original liquid sample by their different ionization propensity under mild electrospray conditions [117, 118]. The possibility to preserve non-covalent interactions, along with conformation-dependent protein ionization, makes it possible to depict conformational ensembles, discriminating conformational states by structural compactness. At the same time, non-covalent complexes are detected by their specific mass, providing a unique possibility to gather combined information on protein folding and binding [119, 120]. Native MS on AS in aqueous solutions detects four main conformational components based on charge-state distributions [121, 122]. These are referred to as compact, extended and intermediate (I and II), by comparison to the reference behavior of folded and unfolded proteins [118].

Native-MS analysis of AS-DA interactions by titration experiments at pH 7.4 [122] provides direct experimental evidence of the formation of non-covalent complexes and shows that no chemical modification of protein or ligand takes place in the time scale of such experiments. DA binds preferentially to a partially structured AS conformer with stoichiometry up to 1:4 and sub-millimolar apparent K${}_D$, and slightly remodels the AS conformational ensemble by stabilizing the intermediate. Ligand footprinting [122] by electron-transfer dissociation (ETD) of the 1:1 complex locates the main interaction site in the ${}_{127}$MPSEE${}_{131}$ region, in agreement with the above reported evidence. The apparent discrepancy with the NMR study failing to detect significant binding at pH 7 [114] could be explained by the peculiar properties of MS, which escapes averaging over the molecular population and, rather, describes folding and binding status of each sorted ion in the spectrum. This property makes MS more similar to single-molecule than to bulk spectroscopy methods [105], offering peculiar sensitivity and deepness in the description of heterogeneous samples. It is interesting to note that, at 1 mM DA and 20 µM AS, ~75% of the 15+ charge state (representing the intermediate state in the native MS spectrum) is bound to DA, while only ~29% of the 8+ charge state (representing the compact state) is bound [122]. In that study, the pH of the DA stock solution was controlled after DA addition [122]. Also relevant to this discussion is the large amount of evidence that rules out uncontrolled pH changes during nano-electrospray of buffered aqueous solutions [123, 124].

Another biophysical method to explore conformational ensembles is single-molecule force spectroscopy (SMFS), in which individual molecules in solution are unfolded mechanically by the cantilever of an atomic force microscopy (AFM) instrument, relating pulling force to the structural features of the molecule [125]. By randomly reiterating this procedure, it is possible to characterize the statistical distribution of distinct protein conformers. Three main conformational states of AS in buffered solutions at pH 7.4 can be detected by this method based on the strength of the intramolecular interactions [105, 126, 127]. The addition of DA shifts the conformational ensemble towards the strong interactions, with loss of the fully unstructured component [105]. These interactions must be at the level of tertiary structure, since the secondary structure content remains unaltered, as discussed above. The structural interpretation of MS and SMFS data has been corroborated by comparison to structural clusters obtained from a 73 µs-long MD simulation of the AS conformational ensemble in solution [128, 129].

AS-DA binding at near-neutral pH has been detected by other orthogonal techniques, such as isothermal titration calorimetry (ITC) and nanopore-permeability analysis. One ITC study at pH 7.8 indicates relatively weak binding, with an estimated dissociation constant of about 280 µM [106]. On the other hand, another ITC study reports too weak binding at pH 7.4 to be quantitatively evaluated [130]. Apparent dissociation constants in the sub-millimolar or millimolar range can be considered biologically relevant, since DA concentrations are of the micromolar order in the cytosol and reach the millimolar range within storage vesicles and striatal nerve terminals [99]. Nanopore-permeability assays provide additional evidence of AS-DA interaction by detecting changes in the frequency distribution of pore translocation or bumping events for the protein in solution, before and after the addition of the ligand [106]. Translocation events are only considered possible for molecules in random-coil conformation, while folded or partially folded structures, instead, bumping away on the cis side of the membrane [131, 132]. According to this interpretation, DA would stabilize a partially folded intermediate in WT AS, while stabilizing the random-coil state in the A30P mutant. Although the results are not of straightforward interpretation in terms of conformational effects, they suggest direct interaction between AS and DA [106]. Thus, the available data suggest that DA itself can bind to AS and affect its aggregation. DA oxidation would not be an absolute requirement, although, as described below, some DA metabolites, such as DA o-quinone and DOPAC, are more effective [97, 100]. Further studies are needed to understand whether apparently conflicting reports in the literature can be interpreted and reconciled by the peculiarities of the different biophysical approaches.

In order to find AS ligands that can inhibit AS fibrillogenesis and toxicity, several Authors have studied the interactions of this protein with DA-related molecules (catabolites, agonists, and analogs). DA analogues are molecules that structurally mimic DA, while DA “agonists” and “antagonists” are molecules able to activate or inactivate the biological response mediated by DA receptors.

One of the first systematic search for direct interactors has been carried out by in-silico screening of 70 compounds and led to the identification of 5 DA analogs that, similarly to 6-aminoindole, are expected to affect the nucleation/polymerization mechanism and possibly to bind oligomers [133] (Fig. 6). In-vitro assays by ThT and AFM has showed that 6-aminoindole and 5-hydroxyindole have the strongest inhibitory effects on AS aggregation, while tyramine and 4-(2-aminoethyl)aniline have the poorest effects [133]. Structural comparison between tyramine and DA suggests that the loss of just one of the -OH groups (replaced by -H in tyramine) causes a dramatic loss in anti-fibrillogenic activity. It has been hypothesized that a benzene ring with two -OH or quinone groups (i.e., 1,2-benzoquinone moieties) may have great potential in AS binding. Stacking interactions and charge interactions appear to be relevant in determining the interaction with either soluble or aggregated AS. Below, DA analogs showing these key structural motifs are discussed. An example of DA analog directly interacting with soluble AS is provided by Fasudil®, an isoquinoline substituted by a (1,4-diazepan-1-yl)sulfonyl group at position 5 (Fig. 6). It also acts as an inhibitor of Rho kinases and serendipitously proved to have antifibrillogenic activity. Its use has been approved by the U.S. Food and Drug Administration (FDA) in clinical trials of disorders such as Raynaud’s disease, atherosclerosis, and age-related or neurodegenerative diseases [134]. Direct interaction of Fasudil® with AS has been primarily demonstrated by NMR [135] and analyzed by MD simulations [136]. These studies offer an atomic-level description of a possible inhibition mechanism. Aromatic stacking and electrostatic interactions are formed through a process termed dynamic shuttling, in which one set of interactions breaks before another is formed nearby [136]. The interactions preferentially involve the C-terminal region of AS, which remains highly dynamic upon binding. Beside the tyrosine-glutamate (YE) residues ${}_{136}$YE${}_{137}$, also the ${}_{125}$YE${}_{126}$ pair appears to be involved in these interactions, although with lower affinity likely depending on the sequence context. The binding of Fasudil® in the C-terminal region may screen protein-protein electrostatic attractions or sterically prevent interactions between monomers.

The category of molecules that bind and dissolve aggregated AS includes three molecules recently isolated by in-vitro high-throughput screening based on ThT fluorescence, light-scattering and transmission electron microscopy (TEM), applied to a library of more than 14,000 compounds [137] (Fig. 6). These are named SynuClean-D [65, 138], ZPD-2 [139] and ZPDm [140], and proved effective even at sub-stoichiometric concentrations in disaggregating mature fibrils of WT AS and familial variants A30P and H50Q. All three compounds are also able to reduce the aggregation of AS in PD models of Caenorhabditis elegans but do not interact with soluble AS, as indicated by NMR studies [140, 141] (Fig. 6). Unfortunately, no structural data are available on the atomic details of the interactions between these disaggregating agents and AS fibrils. Although the three molecules have similar effects, they might have fine differences in their mechanism of action reflecting their structural differences.

For the purpose of drawing some rules about the structure and function of disaggregating molecules, SynuClean-D has been compared with diverse and much more complex compounds, such as EGCG [142] and the molecular tweezer CLR01 [143, 144]. The hypothesis underlying this comparison is that greater structural complexity and larger hydrophobic contact surface (i.e., larger number of aromatic rings) leadto stronger effects in fibril disaggregation [145]. In fact, SynuClean-D, EGCG, and CLR01 contain, respectively, 2, 3, and 5 aromatic rings and this ranking also reflects their disaggregation efficacy (CLR01 $>$ EGCG $>$ SynuClean-D). Nonetheless, it has been observed that small, single-ring molecules can still be highly effective as disaggregating agents [145], while more complex molecules, such as EGCG and CLR01 can have pleiotropic effects combining fibril disaggregation with the ability to directly bind soluble AS and modulate its aggregation pathways, thus reducing the quantity and toxicity of aggregates [146, 147, 148, 149]. This is also reflected in an enhanced ability to attenuate cytotoxic effects in vivo [143, 150]. Furthermore, the effects of polyphenolic molecules may also derive from an indirect protective action against oxidative stresses due to chelation of metal ions [150, 151, 152]. Comparative structural and functional analyses of DA, amphotericin-B, EGCG, and quinacrine dihydrochloride unfortunately confirm our current inability to extract a general model, relating binding affinity and fibril inhibition to chemical structures [150]. The interactions of these compounds to soluble or fibrillar AS, analyzed by NMR and surface plasmon resonance (SPR), show differences in strength and position, with no evident structure-function relationship [150].

DA agonists include molecules that are commonly used to relieve symptoms of PD. Fig. 7 shows the structure of some DA agonists [153, 154]. Among them, the neuroprotectant cabergoline has been suggested to directly interact with AS by molecular docking experiments [155]. Pergolide and bromocriptine have long been recognized by in-vitro ThT assays to also destabilize preformed fibrils by direct interaction [156, 157]. A small group of agonists can be considered bifunctional, in that they also exhibit the ability to modulate AS aggregation by either direct interaction, or by affecting its PTMs and/or turnover [158]. Carbazole-based DA agonists are a recent example of development of such kind of multifunctional drugs [159]. Bifunctional molecules have been designed ad hoc by introducing hydroxyl groups into the structure of known agonists, so that their efficacy on the signaling cascade is not altered and the ability to interact directly with AS is newly conferred. Compounds such as D-519 and D-520 (Fig. 7) prove effective in reducing symptoms in Drosophila melanogaster and rat models of PD, having maintained their efficacy as strong agonists of D2/D3 receptors, and showing new properties derived from their ability to directly interact with AS [160]. In particular, D-520 can reduce the toxicity of preformed AS aggregates against rat pheochromocytoma cell cultures and modulate AS aggregation, as demonstrated in vitro by ThT and TEM [160]. Apomorphine (Apo) is a non-ergoline DA agonist, whose dual function is a downside. Apo binds with high affinity to DA receptors D2, D3, and D5, and can also directly interact with AS [160]. Apo was approved by the FDA in April 2004 for the treatment of hypomobility and dyskinesia associated with PD. Apo inhibits fibrillation and dissolves pre-formed fibrils, as assessed by ThT assays and TEM. Under oxidizing conditions, off-pathway oligomeric adducts, Apo-AS-O, are formed. AS oligomerization is faster in the presence of Apo than DA. Apo-derived AS oligomers are toxic to cultured neuronal cells, and form by recruiting innocuous AS monomers. NMR 1H-15N-HSQC experiments show that Apo binds monomeric AS in the ${}_{125}$YEMPS${}_{129}$ region, as observed for other oxidized catechols. Apo establishes sufficiently stable hydrophobic interactions with monomeric AS to resist dilution during size exclusion chromatography (SEC) [160]. Further studies are needed to understand whether neurotoxic effects of Apo can also occur in vivo [160]. Given that anti-PD drugs can produce toxic intermediates, it is appropriate to reason on their role and the rationale behind their design. Indeed, the formation of specific neurotoxic intermediates, or even the persistence of oligomeric forms derived from fibril disgregation, can have a high neurotoxic potential, and hence should be considered as undesirable effects. Instead, it is desirable to develop drugs that reduce fibrillogenic potential by inhibiting the formation of oligomers, as part of a general preventive and prophylactic strategy to be adopted throughout life, rather than drugs to be used at more advanced stages of a synucleinopathy, when oligomers or fibrils have already formed.

The interaction of AS with DOPAL, the immediate metabolite of DA catabolism, has been proposed as a key factor in PD pathogenesis in the so-called catecholaldehyde hypothesis [32]. The aldehyde group generated by MAO-mediated conversion of the DA amine group makes DOPAL capable of quinonizing AS, forming covalent adducts to lysine residues or to the N-terminus by Schiff base or Michael addition, or generating a dicatechol pyrrole lysine structure [161]. ThT fluorescence experiments have shown that DOPAL destabilizes preformed AS fibrils and inhibits aggregation, generating off-pathway oligomers [162]. According to near-infrared fluorescence spectroscopy and western-blotting experiments, DOPAL induces the formation of AS oligomers more effectively than DA [163]. DOPAL-induced AS quinonization and oligomerization are enhanced in the presence of Cu(II) ions [163]. Despite some contrasting evidence [164], a consensus is emerging in the literature that many DOPAL-induced oligomers are toxic and compromise several physiological processes in neurons [165]. Blocking the DOPAL aldehyde moiety by compounds containing primary (e.g., aminoindan) or secondary (e.g., rasagiline) amines abrogates its effects on protein aggregation and cytotoxicity [165].

AS oligomers generated by DOPAL have been investigated by SEC-HPLC [166]. Detected species have been classified, according to their size, in small (dimers and trimers) and large (tens of subunits) oligomers. The relative amounts of these species are finely tuned by several factors (mutations, PTMs, etc.). Among them, DOPAL-induced oxidation of methionine residues has been shown to favor the accumulation of small oligomers, hampering the formation of larger species [166, 167]. In particular, the oxidation of the C-terminal methionine at position 127 improves the Ability of AS to scavenge ROS derived from DOPAL and prevents the assembly of large neurotoxic oligomers [167].

DOPAL-mediated oligomerization is also influenced by familial PD point mutations of AS. In particular, the A53T, E46K and H50Q variants of AS have been shown to stimulate formation of large (n $>$ 3) rather than small (n = 2–3) oligomers upon the incubation with DOPAL [163, 168]. DOPAL-induced large oligomers have been associated with enhanced toxicity [166]. N-terminal acetylation and binding to phospholipid vesicles counteracts DOPAL-induced oligomerization, as shown by denaturing electrophoresis and SEC experiments [168].

The DA metabolite DOPAC is the product of DOPAL detoxification by aldehyde dehydrogenase conversion, and it is usually rapidly removed by active transport from the cell via a sulfonylurea-sensitive transporter [169]. As its precursor DOPAL, it can undergo spontaneous oxidation to a quinone, thus producing H${}_2$O${}_2$, a reactive molecule able to disturb the redox balance [170].

It has been observed in vitro that DOPAC at physiologically relevant concentrations binds non-covalently to AS and stabilizes the oligomeric state, hindering fibrillogenesis [170]. Because of the oxidative instability of the compound, the molecular mechanism could not be unveiled, but this effect might be due to an alteration of the oligomer interface that plays a fundamental role in fibril formation. The interaction has been mapped by mutagenesis to the same N-terminal region that binds lipid vesicles [163]. At high molar excess (i.e., DOPAC:AS $>$3:1), DOPAC forms covalent adducts with AS, as also seen for DOPAL [170]. A following study demonstrated that DOPAC binding also decreases the ability of AS to interact with lipids, thus reducing its functionality [171]. Furthermore, AS/DOPAC oligomers showed a reduced propensity to conformational changes upon interaction with cellular membranes [172]. DOPAC-AS interaction has an influence also on the redox state of both molecules. Indeed, AS speeds up the spontaneous oxidation of DOPAC, whereas the oxidized form of DOPAC (the quinone derivative) leads to oxidation of AS methionine residues, probably due to H${}_2$O${}_2$ generation, as monitored by MS along an in-vitro kinetic assay [170]. AS oxidation has been reported to promote the formation of oligomers, rather than fibrils, in vitro [173]. The addition of catalase as an H${}_2$O${}_2$ scavenger prevents AS oxidation but does not affect AS aggregation properties, as assessed by ThT, TEM and SEC. These results suggest that the oxidation of AS methionines is not required for oligomerization and inhibition of fibrillogenesis by DOPAC [172].

As discussed in the previous section, DOPAL is further oxidized by mitochondrial aldehyde dehydrogenase to DOPAC, in the major oxidation pathway of DA (Fig. 2). In a minor pathway (under normal conditions), DOPAL can be, instead, reduced by alcohol dehydrogenase (ADH) to 3,4-dihydroxyphenylethanol (DOPET), also called hydroxytyrosol (Fig. 2). In addition to this endogenous source as DA metabolite, DOPET can have a dietary origin, being found in olive oil and wine. Moreover, DOPET can be produced from the hydrolysis of olive oil secoiridoids, such as oleuropein. Noteworthy, DOPET is able to cross the blood-brain barrier and exhibits antioxidant, anti-inflammatory, and anti-aggregation activities [174].

In a first in-vitro study [175], DOPET was found to inhibit AS aggregation and to destabilize preformed AS fibrils, as evaluated by ThT and TEM investigations. In particular, DOPET inhibits AS fibrillogenesis, leading to the formation of small amorphous aggregates. Moreover, DOPET was found to counteract AS-induced toxicity on PC12 cells [175].

The effects of DOPET on AS aggregation have been elucidated by employing complementary biochemical and biophysical approaches, including ThT assay, CD spectroscopy, TEM, native MS, limited proteolysis, intrinsic fluorescence, SEC and electrophoretic analyses [176]. ThT assay and TEM show that a 1:1 DOPET:AS ratio completely inhibits AS fibrillogenesis, leading to the formation of AS oligomers, which appear structurally disordered as indicated by CD spectroscopy. The CD analyses also suggested that DOPET does not affect the secondary structures of monomeric AS. Chromatographic, TEM, and MS analyses of AS treated with DOPET showed the presence of spherical oligomers and that of protein species containing one-to-four oxidations of methionine residues, whose speed of formation and extent of oxidation were related to DOPET concentrations. Covalent adducts, likely involving the $\epsilon$-amino group of lysine residues and the aromatic ring of DOPET, were also observed. Limited-proteolysis experiments have revealed similar fragment patterns at the beginning of the incubation with and without DOPET [176]. After longer incubation, AS alone displays a reduced susceptibility to proteolysis, in agreement with the formation of AS fibrils, while the protein species formed in the presence of DOPET are more susceptible to proteolysis. A slight protection to proteolysis of the C-terminal region was observed in the monomeric protein in the presence of DOPET, suggesting that this region could be involved in the AS-DOPET interaction [176]. The effects of DOPET on AS aggregation were also studied in the presence of catalase, to inhibit the methionine oxidation induced by DOPET. DOPET retained the anti-fibrillogenic activity also in the absence of AS oxidation [176]. The presence of DOPET during AS aggregation remarkably decreases AS cytotoxicity on SH-SY5Y human neuroblastoma cells, as measured by the MTT test. Since aggregated AS enhances intracellular ROS and binds the cell membranes at monosialotetrahexosylganglioside 1 (GM1)-enriched sites, both events were tested in the presence of DOPET. DOPET-induced AS aggregates displayed reduced ROS and membrane GM1 binding, in comparison to AS aggregates obtained in the absence of the compound. Overall, this study showed that DOPET interacts with AS by non-covalent and covalent interactions and redirects protein aggregation to the formation of not-toxic spherical oligomers and amorphous aggregates, which appear to be off-pathway of fibrils formation [176].

The interactions of DOPET and DOPAC with WT AS and the E46K pathological mutant have been compared [177]. Both compounds partially or completely inhibit fibril formation at 1:1 or 1:5 protein:catechol ratio, respectively. In the absence of catechol, WT and E46K formed fibrils with a different morphology, as observed by TEM. DOPAC at 1:5 protein:catechol ratio induced the formation of spherical and annular oligomers of 10–20 nm and 20–40 nm for E46K and WT AS, respectively. In the presence of DOPET, mainly amorphous structures were observed under the tested experimental conditions. The formation of oligomeric species was also investigated by SEC, which showed additional peaks due to protein dimers and trimers in the presence of each compound. A higher fraction of oligomeric species was observed in the early stages of incubation for E46K, both in the absence and the presence of catechols, compared to WT AS. CD spectroscopy showed that the off-pathway species induced by DOPET and DOPAC are characterized by a predominant random-coil structure, similar to monomeric AS. In all the above investigations, DOPET appeared less active in affecting AS fibrillogenesis compared to DOPAC and the E46K mutant appeared less affected by catechols compared to the WT AS [177].

Important structural details of the catechol effects on AS fibrillogenesis were obtained by SEC, native-MS and hydrogen-deuterium isotope exchange (HDX)-MS. At the beginning of the incubation in the absence of catechols, E46K eluted at a slightly higher volume in SEC compared to WT AS suggesting increased compactness of the mutant. In the presence of DOPAC or DOPET, E46K eluted slightly earlier, suggesting a stabilization of a more extended conformer. On the contrary, catechols did not affect the retention time of the WT protein. Native-MS indicated the presence of multiple protein conformers with different compactness at the beginning of incubation, in agreement with other studies [177]. The addition of DOPAC shifted the equilibrium towards a more extended conformer, an effect that was more evident for the WT protein. Compared to DOPAC, the effect of DOPET was less evident on E46K and not significant in the case of WT AS. The most remarkable effects of both catechols were observed by native-MS at 48 hours of incubation for E46K and WT AS [177]. ThT assays and TEM indicated that DOPAC, and DOPET to a minor extent, were able to disaggregate intermediates and mature E46K aggregates obtained in the absence of the compounds. This effect was less evident in the case of mature WT AS aggregates.

On the basis of these results [176, 177], the Authors proposed the following model for the effects of the two catechols on AS aggregation. Catechol binding affects the conformational equilibrium of monomeric AS, stabilizing a conformer that is less prone to convert into amyloid fibrils. These extended monomers assemble into not-toxic, off-pathway oligomers and amorphous aggregates. AS oxidation is not required for DOPET inhibition of AS fibril formation [176, 177].