Pre-aggregation kinetics and intermediates of α-synuclein monitored by the ESIPT probe 7MFE

The defining feature of the extensive family of amyloid diseases is the formation of networks of entangled elongated protein fibrils and amorphous aggregates exhibiting crossed β-sheet secondary structure. The time course of amyloid conversion has been studied extensively in vitro with the proteins involved in the neurodegenerative pathology of Parkinson’s disease (α-synuclein), Alzheimer’s disease (Tau) and Huntington’s disease (Huntingtin). Although much is known about the thermodynamics and kinetics of the transition from a soluble, intrinsically disordered monomer to the fibrillar end state, the putative oligomeric intermediates, currently considered to be the major initiators of cellular toxicity, are as yet poorly defined. We have detected and characterized amyloid precursors by monitoring AS aggregation with ESIPT (excited state intramolecular protein transfer) probes, one of which, 7MFE [7-(3-maleimido-N-propanamide)-2-(4-diethyaminophenyl)-3-hydroxychromone], is introduced here and compared with a related compound, 6MFC, used previously. A series of 140 spectra for sparsely labeled AS was acquired during the course of aggregation, and resolved into the relative contributions (spectra, intensities) of discrete molecular species including the monomeric, fibrillar, and ensemble of intermediate forms. Based on these findings, a kinetic scheme was devised to simulate progress curves as a function of key parameters. An essential feature of the model, one not previously invoked in schemes of amyloid aggregation, is the catalysis of molecular fuzziness by discrete colloidal nanoparticles arising spontaneously via monomer condensation upon exposure of AS to ≥ 37 °C. Electronic supplementary material The online version of this article (10.1007/s00249-017-1272-0) contains supplementary material, which is available to authorized users.

Procedure. Take a 250 ml round flask and add 53 g of dry K 2 CO 3 (384 mmol) and a stirring bar. Then place flask in an oil or sand-bath, connect a reflux condenser and purge the system with Argon. Then add 50 ml of dry acetone and mix vigorously, creating a suspension. Add 10.0 g (66.2 mmol) of N-(3-hydroxyphenyl)acetamide (25) and 19 ml of methyl sulfate (200 mmol, d=1.33 g/ml). Both are available from Sigma Aldrich. After that, the system is heated and allowed to reflux for 16 h. The following day, heating is turned off and the reaction is allowed to cool to RT. The suspension is the filtered and washed with acetone. The flow-through is tested after each wash by depositing a drop on a TLC silica gel plate and examining under UV. The solid is washed until no visible spot appears under UV. The acetone solution is saved and the solid discarded. The excess acetone is evaporated in a rotary evaporator. As acetone is removed and the solution becomes concentrated, a white precipitate becomes visible. Acetone is not completely removed, and enough is added so as to redissolve the white precipitate. A 5% solution of NH 3 in water is added slowly to basify and help destroy any leftover methyl sulfate. DCM or chloroform are then used to extract the desired compound from the aqueous basic solution. The organic phase is saved, dried with sodium sulfate (anh.), filtered and evaporated.
Procedure. Add 30 ml of dried carbon disulfide (CS 2 ) in a round flask. Place the flask in a water/ice bath and allow 10-15 min to equilibrate. Add 4.0 g (24.3 mmol) of compound 26 (obtained from step 1). Mix vigorously. Let the suspension cool, then, add 5.0 ml (70 mmol) of acyl chloride, followed by aluminum trichloride (12 g, 90 mmol). AlCl 3 should be added in 4-5 small portions over a period of 30 min so that the temperature does not rise above 35-40 °C after each addition. When all AlCl 3 has been added, the reaction is allowed to reach RT over an hour and then heated until reflux over 3 h. After this reflux time, the reaction is cooled to RT and then poured into water/ice where the product precipitates.
Procedure. In a 1 L round flask add 752 mg (3.89 mmol) of compound 27 (obtained in step 2) and 750 mg (4.24 mmol) of 4-(dimethylamino)benzaldehyde (20), avaikble from Sigma Aldrich, and 10 ml of dry DMF. Dissolve everything. Add 750 mg (13.89 mmol) of sodium methoxide and leave stirring ON at RT. The next day, prepare a heating mantle and a bucket full of ice. It is important to have this ready before starting by adding 30 ml of ethanol to dilute the reaction mixture and facilitate boiling more. Briefly mix well, then, add 5 g (92.6 mmol) of sodium methoxide. Briefly mix the suspension. Finally add all at once 7 ml of 30% hydrogen peroxide. The solution should start to heat up and boil almost immediately. If it doesn't , briefly place it on a pre-heated mantle to help the flask reach the appropriate temperature. When the reaction mixture boils (by observation of bubbles and/or condensation of ethanol on the side or top of the flask), remove from mantle. The reaction mixture should be able to self-sustain the temperature. Allow it to boil for only 3 min (no more than 5 min). Then place and bury flask on ice and cool it down for ∼15 min. Finally pour the reaction mixture on ice/water in a 500 ml flask or beaker provided with a stirring bar. Neutralize to pH 5-6 with HCl, checking with pH-paper strips. A color change should be evident in the solution while neutralizing. At a pH of 5-6, the solution becomes a brown fine suspension. If the pH drops below the indicated limit, the suspension changes again into a solution. Regardless the pH, when a very fine suspension is noticeable stop and filter using an appropriate paper (for fine solid)s.

Purification.
The desired compound is purified by normal column chromatography using cyclohexane/ethyl acetate (from 9:1 to 7:3), yielding 568 mg (1.55 mmol, 40%) of an orange(-ish) product 28. Before purification, the product should be checked by TLC. A yellow-orange spot should be apparent, with fluorescence under long wavelength illumination (~365 nm).

MS (ESI
Procedure. In a round flask, add 200 mg (0.546 mmol) of product 28 (obtained in step 3) in 5 ml of 10% HCl (aq) with a 15% of ethanol. Heat to reflux for 3-5 h. Then add some water and neutralize to pH 6-7. An orange (fine) solid should precipitate. Filter, wash with small amount of water, and dry under vacuum ON. This can be use without further purification (checked by TLC and NMR), yielding 150 mg (0.462 mmol, 85%) of product 29. By TLC this should also fluoresce under long wavelength illumination.
Alternative procedure. Although the following procedure was not explicitly tested for these compounds, it was tested for amide isomer. If available, this reaction is possible under Microwave irradiation instead of heating reflux. Example: in a 10 ml microwave vial, 23 mg of the acetamide-chromone is suspended in 3 ml of 10% HCl plus 1 ml of 99% ethanol. The suspension is sonicated briefly to achieve maximum dispersion. The vial is sealed and heated to 140-150 °C for 15min with 500 rpm mixing (Anton Paar Monowave 300). Proceed to filter, wash with water and dry under vacuum ON. Yield: 85-95% Step 5. Synthesis of 7-MFE (30).
Procedure. In a round flask, dissolve 15 mg (0.046 mmol) of product 29 (obtained in step 4) in 5 ml anhydrous THF. Add 3-maleimidepropanoic acid (11, 7.8 mg, 0.047 mmol). Then add 500 µl of DIC (09) at RT. After 3h the reaction is filtered and carefully washed with DCM. Purification: the washed solid is purified by preparative TLC using cyclohexane/ethyl acetate (6:4) and revealed by both short and long UV wavelength. The scratch band is then washed and product extracted with Ethyl Acetate, obtaining product 30 with a 35% yield (7.6 mg, 0.016 mmol).

AS fluorescent labeling.
Mono-cysteine mutants of AS were covalently labeled with MFC and 7MFE as described before in (Fauerbach et al., 2012). For the labeling procedure, the maleimide fluorophores (MFC & 7MFE) were dissolved in DMSO and mixed dropwise with a solution of ∼300 µM of monocysteine AS (i.e. AS A18C, AS A140C) in 25 mM Na-PO 4 pH 7.3 with 3 mM tris(2-carboxyethyl)phosphine (TCEP); the final DMSO:buffer ratio was 1:2. MFC was in a 5-10-fold molar excess over the protein. The solution was left at 4 ºC under mixing overnight (ON). The next day the reaction mixture was diluted 10-fold with 25 mM Na-PO4 pH 7.3, 3 mM TCEP and concentrated by 3-4 passages through a Millipore Amicon 10 kDa filter to a volume of ~ 1 ml in order to remove DMSO from the reaction mixture. A size exclusion PD-10 column was used as a first step of purification from the unbound dye. A Pharmacia Smart chromatography system was used with a Superdex 200 column (10/300) and 25 mM Tris-HCl, pH 7.2, 150 mM NaCl elution buffer to separate thoroughly the excess reagent from the labeled protein and the unlabeled protein. The extent of labeling was 50-70%. Samples were frozen with liquid N 2 and stored at -80 ºC.

AS aggregation.
Aggregation assays of wtAS protein in combination with cysteine-containing mutants labeled with the MFC ESIPT probe were performed at 37 °C, as described previously (Fauerbach et al., 2012, Yushchenko et al., 2010.
6. Spectral deconvolution. The procedure is described in the text. All programs and calculations were implemented in Mathematica 11.1. Global fits of a set of hyperbolic functions to the initial and final phases of the aggregation reaction served to define sp1 and sp3. For example, the functions used for the 7MFE data were as follows.  Scan 140 + fit The differential equations corresponding to this scheme are: where m = monomer, m1 = cluster, m2 = colloid, m3 = fuzzy, m4 = nucleus, and m5 = fibrils, and m0 = total monomer concentration. Cluster, colloid, fuzzy and nucleus are regarded as nth order condensations of monomer; in the equations, the respective order parameters are n1, n2, p, and p. Simulations are performed typically with n1 = 10, n2 = 2, and p = 2. In the equations k1a = k1' and k2a = k2', where k1' and k2' are identified in Figure 8.
Convenient estimations of k1 and k1a are obtained from the formulation for the relaxation time τ of an isolated concerted nth order association (Bernasconi, 1976), often invoked for describing the nucleation of fibrillation (Saric, Michaels et al., 2016).
A preliminary assessment of the dependence of important reaction features in the individual parameters of the equations is given in Table S1. Examples of the complex influence of the key rate constant k3 are featured on Page 8. Table S1. Influence of each rate constant on the performance of the scheme, Preliminary.