Disassembly of preformed amyloid beta fibrils by small organofluorine molecules

Abstract

A potential therapeutic approach for Alzheimer’s disease is to reduce the amount of toxic amyloid-beta oligomers and fibrillar amyloid plaques. In order to contribute to this approach the ability of small organofluorine compounds that were previously reported as successful inhibitors of fibrillogenesis to destabilize preformed fibrils of the amyloid-beta peptide was studied. These organofluorine molecules including chiral compounds were tested in vitro using standard methods based on Thioflavin-T (THT) fluorescence spectroscopy, atomic force microscopy (AFM) and Fourier-transform infrared spectroscopy (FTIR). It was observed that 5′-halogen substituted 3,3,3-trifluoromethyl-2-hydroxyl-(indol-3-yl)-propionic acid esters showed significant activity in the disassembly of the preformed fibrils. Since the same compounds were identified as strong fibrillogenesis inhibitors as well, this dual action makes them promising candidates for further drug development.

Alzheimer’s disease (AD) involves misfolding and aggregation of the amyloid-beta peptide (Aβ) and its deposition into characteristic fibrils called amyloid plaques.¹ In the traditional amyloid cascade hypothesis of AD these fibrils are associated with pathology; while recent studies also emphasize the highly neurotoxic nature of precursor oligomers, which, in fact, are suggested being the most important link to neurotoxicity.^2,3,4 Therefore, the inhibition of the seeding and progress of Aβ self-assembly, complimented by the disassembly of the preformed aggregates is among the promising therapeutic options. One of the most pursued strategies is the inhibition of Aβ oligomerization and/or fibrillization by small organic molecules.^5,6 Some of these studies were also extended to the reverse process; the destabilization of the preformed fibrillar aggregates.^5,6,7,8 The modes of action of these molecules are still unclear, mostly due to the limited structural information available on the diverse amyloid assemblies.^9,10 The structure-activity relationship studies using strategic chemical modifications of the effectors can reveal important interactions and conformational requirements between ligands and peptides and also between the associating peptide units. Therefore, these studies can significantly contribute to the general understanding of amyloid formation and lead to a more rational design of new drug candidates^{11, 12}.

Continuing our earlier efforts on designing novel inhibitors of Aβ self-assembly, herein we describe the ability of small organofluorine inhibitors to act as disassembly agents. We have found that a number of trifluoromethyl-hydroxyl-(indol-3-yl)-propionic acid ethyl esters strongly inhibited the in vitro formation of Aβ fibrils.^13,14 In the present work we extend our above studies and determine the effect of these inhibitors on the stability of preformed fibrils. The main goal is to identify compounds that are not only able to inhibit, but also to reverse the fibril formation. The chemical structure of the organofluorine compounds that have been applied in this study are shown in Fig. 1 together with melatonin, a well-known amyloid inhibitor,¹⁵ used for comparison.

Figure 1.

(a) Organofluorine molecules used in the disassembly studies. (b) Melatonin, a well-known amyloid inhibitor, was also studied for comparison.

The proposed compounds were synthesized and obtained in excellent yields and selectivities using procedures described previously.¹⁶ In addition to the racemic compounds, three enantiomer pairs have been synthesized, as chiral compounds may have significantly different effects on biomolecules. The structures of the chiral molecules tested are shown in Fig 2.

Figure 2.

Structure of the enantiomeric trifluoro-hydroxy-(indol-3yl)-propionic acid esters.

For the disassembly experiments the fibrils were grown following the same protocol as described earlier.^{13,14, 17}, ^† The data obtained after 4 days of incubation are shown in Fig. 3. They clearly indicate the ability of these racemic as well as enantiomeric organoflourine molecules to disassemble the preformed fibrils although to different extent. The best results were obtained with the iodo-derivatives 9, 9(R), 9(S). Furthermore, only negligible differences were found for the enantiomeric pairs indicating little or no role of chirality in the disassembly. Both of these results are similar to our earlier findings in case of fibril inhibition.¹⁴

Figure 3.

Relative THT fluorescence in the presence of (a) racemic trifluoro-hydroxyl-(indol-3-yl)-propionic acid esters and melatonin; (b) chiral 5′-halo derivatives of trifluoro-hydroxyl-(indol-3-yl)-propionic acid esters at 10:1 (small molecule:Aβ) molar ratio. The measured fluorescence intensities (I_THT) were normalized to the control sample C containing Aβ only (relative fluorescence: 100%). The data are average of 3 parallel experiments for (a) and 4 parallel experiments for (b), the values represent means ± standard deviation.

Morphological characterization of the samples was carried out by Atomic Force Microscopy, using a Quesant Q-Scope 250 microscope in non-contact mode following the procedure described previously.^14,18, ^‡ The images are shown in Fig. 4.

Figure 4.

Atomic Force Microscopy images of Aβ_1-40 samples after the incubation of the preformed fibrils with the test compounds for the given time (a) C (control)-4 days, 10 μm² scan with z-axis of 87.59 nm, (b) M-4 days, 6 μm² scan with z-axis of 130.8 nm (c) 9(S)-4 days, 10 μm² scan with z-axis of 16.01nm, (d) 7(S)-4 days, 10 μm² scan with z-axis of 77.40nm, (e) 9(R)-0 day, 5 μm² scan with z-axis height of 69.13 nm, (f) 9(R)-1 day, 3 μm² scan with z-axis height of 57.89 nm, (g) 9(R)- 4 days, 10 μm² scan with z-axis of 342.5nm, (h) 9(R)- 6 days, 5 μm² scan with z-axis height of 27.53 nm.

The AFM images correlate well with the fluorescence data. The image of the control shows well-developed fibrils as expected (Fig. 4a). There is no significant morphological change in the presence of melatonin either (Fig. 4b). However, samples incubated with the small organofluorine compounds, clearly showed disaggregated fibrils. The process of disaggregation could be observed unambiguously in the case of iodo-derivatives (9(S) Fig. 4c, and 9(R) Fig. 4e-h) as well as 7(S) (chloro-derivative, Fig. 4d). The efficacy order in which the molecules disassembled the fibrils, irrespective of the stereospecificity, is 9, 8 and 7. Fig. 4e-h illustrates the disassembly process as a function of time. The images depict the gradual disassembly by 9(R) from the starting point (0 day Fig. 4e) through day 1 (Fig. 4f) and day 4 (Fig. 4g) to day 6 (Fig. 4h).

Structural changes due to the disassembly of fibrils were also followed by Fourier-transform infrared spectroscopy (FT-IR) focusing on the amide I (~1650 cm⁻¹) region of the absorption spectra (Fig. 5.). The amide I vibrations are mainly attributed to the C=O stretches in the polypeptide backbone and most commonly used to perform secondary structure analysis.¹⁹,^§ Spectra obtained on day 0 of disassembly were dominated by an amide I vibration at 1630 cm⁻¹ characteristic of β-sheet rich secondary structures (C₀ and C₄ in Figs. 5a and 5b).¹⁹ The frequency of the dominating band is shifted from 1630 cm⁻¹ to around 1645-1650 cm⁻¹ indicating the formation of oligomers or other less ordered species. In agreement with the THT fluorescence and AFM results, the FT-IR data also indicated that the iodo-derivatives were the most efficient in disassembling the fibrils as observed by the shifts in the peak frequencies (9₀ and 9₄ in Fig. 5a). The appearance of a broad featureless band at around 1650 cm⁻¹ was attributed to random coil structures.¹⁸

Figure 5.

The absorbance FTIR spectra (showing the characteristic amide I region) obtained on the day 0 (dotted line) and on the day 4 (solid line) of the disassembly experiments in the presence of (a) selected racemic organofluorine compounds and melatonin (b) selected chiral organofluorine compounds. The label C indicates the control Aβ samples with no affector molecules added.

Other organofluorine derivatives studied were also able to disassemble the fibrils to some extent (7₀ and 7₄, 8₀ and 8₄ in Fig. 5a), and this property was independent on the chirality (7(R), 7(S), 8(R), 8(S), 9(R), 9(S) in Fig. 5b). Melatonin, which is a known inhibitor, was used as reference and it was observed that despite its good inhibiting potential, it was not able to disassemble the preformed fibrils (M₀ and M₄ in Fig. 5a).

The concordant data obtained by the various analytical methods clearly support the conclusion that 5′-halogen substituted 3,3,3-trifluoromethyl-2-hydroxyl-(indol-3-yl)-propionic acid esters tested above have a good potential for disassembling the mature fibrils, and this ability is irrespective of the stereospecificity associated with each compound. Since the same molecules were also good fibrillogenesis inhibitors,^13,14 these compounds can be classified as ‘dual nature inhibitors’. This dual action makes them promising candidates for further drug development.