Foldamer-Mediated Structural Rearrangement Attenuates AÎ²

Subscriber access provided by READING UNIV

Article

Foldamer-mediated Structural Rearrangement Attenuates A# Oligomerization and Cytotoxicity Sunil Kumar, Anja Henning-Knechtel, Ibrahim Chehade, Mazin Magzoub, and Andrew D. Hamilton J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08259 • Publication Date (Web): 23 Oct 2017 Downloaded from http://pubs.acs.org on October 23, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Foldamer-assisted Conformational Induction in Aβ

Kumar et al

Foldamer-mediated Structural Rearrangement Attenuates Aβ Oligomerization and Cytotoxicity Sunil Kumar1*, Anja Henning-Knechtel2, Ibrahim Chehade2, Mazin Magzoub2, and Andrew D. Hamilton1*

*Correspondence: [email protected], [email protected] Addresses: 1Department of Chemistry, New York University, New York, NY 10003, USA. 2

Biology Program, Division of Science, New York University Abu Dhabi, Abu Dhabi, United Arab Emirates.

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 2 of 33

Kumar et al

Abstract The conversion of the native random coil amyloid beta (Aβ) into amyloid fibers is thought to be a key event in the progression of Alzheimer’s disease (AD). A significant body of evidence suggests that the highly dynamic Aβ oligomers are the main causal agent associated with the onset of AD. Among many potential therapeutic approaches, one is the modulation of Aβ conformation into off-pathway structures to avoid the formation of the putative neurotoxic Aβ oligomers. A library of oligoquinolines was screened to identify antagonists of Aβ oligomerization, amyloid formation, and cytotoxicity. A dianionic tetraquinoline, denoted as 5, was one of the most potent antagonists of Aβ fibrillation. Biophysical assays including amyloid kinetics, Dot Blot, ELISA, and TEM show that 5 effectively inhibits both Aβ oligomerization and fibrillation. The antagonist activity of 5 towards Aβ aggregation diminishes with sequence and positional changes in the surface functionalities. 5 binds to the central discordant α-helical region and induces a unique α-helical conformation in Aβ. Interestingly, 5 adjusts its conformation to optimize the antagonist activity against Aβ. 5 effectively rescues neuroblastoma cells from Aβ-mediated cytotoxicity and antagonizes fibrillation and cytotoxicity pathways of secondary nucleation induced by seeding. 5 is also equally effective in inhibiting preformed oligomer-mediated processes. Collectively, 5 induces strong secondary structure in Aβ and inhibits its functions including oligomerization, fibrillation, and cytotoxicity.

2


Page 3 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Kumar et al

Introduction Various foldamers have been designed to mimic the shapes and surface properties of different biopolymers including proteins, nucleic acids, and polysaccharides1-5. Foldamers are generally described by their intrinsic capacity to fold into well-defined conformations stabilized by a number of non-covalent interactions2-4. A wide range of foldamer structures have been reported but particular focus has been placed on iterative amide bond formation due to easy availability of the monomers and the potential for the formation of inter-amide hydrogen bonding (Figure 1a,c)2,

3, 6-10

. Oligoarylamides have the capacity to form various secondary structures, from

helical to linear strand8,

10-12

which allow them to target various biomolecules. Thus

oligoarylamides represent a new class of potential biomodulators the important work of Huc and Gong in advancing this field

13, 14

7, 10, 11, 13

and of particular note is

. Our interest has been in

using oligoarylamides (Figure 1d) containing different functionalities on their surface to modulate the sometimes lethal biochemical properties of intrinsically disordered proteins such as islet amyloid polypeptide (IAPP) or Aβ6,

15-18

. We have previously reported the inhibitory

behavior of rigid-rod shaped pyridylamide foldamers (Figure 1d) to modulate the aggregation of IAPP15-18, which is associated with the onset and progression of type II diabetes. In this paper we exploit the elegant quinoline scaffold of Huc to disrupt the aggregation and cytotoxicity of Aβ19. Alzheimer’s disease (AD), the most common type of neurodegenerative disorder, is characterized by an accumulation of soluble proteins and their conversion into intractable amyloid fibrils in the central nervous system20-23. A central event that underlines the etiology of AD is the aggregation of Aβ4224, 25. Numerous research efforts have demonstrated that the soluble Aβ oligomers are the key neurotoxic species that are associated with the progression of AD24, 26. Therefore, a strategy that intervenes with the formation of soluble Aβ oligomers represents a potential therapeutic approach for the treatment of AD. A plethora of ligand based approaches have been employed to interfere with Aβ oligomerization and the associated downstream toxic functions, including (-) epigallocatechin-3-gallate (EGCG) and related polyols,27-29 resveratrol,30 aromatic foldamers,31 molecular tweezers (MTs),32 quinoline-, pyridine-, and pyrrole-based chemical regulators,33 phenol-triazole ligands,34 chaperones,35 ligand D-737 and its derivatives,36 cucurbit[7]uril,37 β-cyclodextrins,38 cyclic peptides,39-41 Aβ-based short peptide domains,42 dyes,43,

44

curcumin,44,

45

aminopyrazole and its derivatives,46 N-methyl peptides,47 β-sheet 3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 33


Kumar et al

breakers48, and Aβ-sheet mimics49, 50. Although these approaches are very effective in inhibiting Aβ aggregation and providing significant mechanistic insights into Aβ fibrillation, some of them are beset by serious limitations. For instance, some of these ligands are non-specific inhibitors of Aβ, while others block the later stages of Aβ aggregation, which will favor the formation of preamyloid toxic oligomers of Aβ. In an alternate approach, ligands were designed to bind and induce a conformational change in Aβ in order to alter its downstream toxic functions. Examples include constraining the conformation of Aβ to an α-helix (Figure 1e) or a β-sheet by antibody/peptidomimetics6,

51, 52

or antibody/protein affibodies53,

54

, respectively. The latter

approach is particularly effective as the affibody adapts to the conformational preferences of its target by modifying its conformation, thereby maximize the binding interaction. In this report, we describe a related approach in which a dynamic foldamer interacts and induces an α-helical conformation within Aβ and itself adopts a right-handed helical conformation to maximize the binding interaction. The screening of a series of oligoquinolines using an amyloid aggregation assay led to the identification of a dianionic tetraquinoline, 5, as one of the most potent antagonists of Aβ aggregation. Immunoassays indicate that 5 suppresses oligomeric intermediates of Aβ. Circular dichroism (CD) and heteronuclear single quantum correlation (HSQC) NMR experiments suggest that 5 induces strong α-helical structure in Aβ and that the binding site resides in the central α-helical domain of Aβ. Fluorescence titration between Aβ and 5 yielded an apparent Kd in the sub-micromolar range. 5 rescues mouse neuroblastoma cells from Aβ induced cytotoxicity and was equally impressive in inhibiting secondary nucleation processes and preformed oligomers of Aβ.

4


Page 5 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Kumar et al

Results and Discussion The Design of Oligoquinoline-based Foldamers Numerous reports have suggested that amyloidogenic proteins can cross-seed each other and augment amyloid formation55, 56. IAPP and Aβ share a sequence similarity of ~50% with ~25% identical positioning of the amino acid residues. For instance, Aβ(15-21) and Aβ(26-32) sequences share a high degree of sequence similarity and commonality with IAPP(10-16) and IAPP(21-27)55,

57

, respectively, and are thought to provide the core nucleation site for

amyloidogenesis. This suggests that small molecules that are effective antagonists of the aggregation of one peptide may also be antagonists for the other57. Miranker et al have shown that oliquinoline-based foldamers (Figure 1a,c) effectively inhibit IAPP aggregation under both solution and cellular conditions1, 58-60. We reasoned that the oligoquinolines would be effective antagonists of Aβ aggregation and its downstream processes and tested a library of their derivatives against Aβ aggregation. Antagonism of Aβ Amyloidogenesis by Foldamers A ThT-based amyloid assay61 was employed to monitor the aggregation of Aβ, resulting in a sigmoidal curve with a t50 of 2.5±0.2 h and 28±4 h for Aβ42 (5 µM) and Aβ40 (25 µM), respectively (Figure 2a; Figure S1). Among the oligoquinolines, 5 was the most potent antagonist of Aβ42 and Aβ40 aggregation (Figure 2 a,d). The aggregation of Aβ was wholly suppressed in the presence of 5 at an equimolar ratio (Figure 2 a,d) while at a sub-stoichiometric ratio of 1:0.5 (Aβ42:5), the ThT intensity was attenuated by ~70% (Figure 2d). Structural modifications of 5 that replaced the 2-methyl butyl group with charged carboxylate group or other hydrophobic groups (ethyl, isopropyl, and isobutyl groups) were detrimental to the antagonism of Aβ aggregation (Figure 1a; Figure S2 for chemical structures). Positional mutations in the side chain functionalities diminished the inhibitory activity of the oligoquinolines (Figure 2c), as seen in the diminution of the fibrils of Aβ42 from 100% to 5%, 50%, and 45% in the presence of an equimolar amounts of 5, 6, and 7, respectively (Figure 2c). 5 does not self-aggregate under the experimental conditions used in various assays as confirmed by TEM, UV spectroscopy, and dynamic light scattering (data not shown). In comparison to 5, the natural products, EGCG27 and Curcumin45, were moderately effective in delaying Aβ 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 6 of 33

Kumar et al

aggregation with t50’s ~2 fold greater than the control reaction (Figure 2b). The mutation study (functional and positional) suggest that the origin of Aβ amyloid inhibition activity is specific and governed by the functional groups presented on the surface of 5. Immunoassays and TEM show that 5 inhibits the formation of the neurotoxic Aβ oligomers. TEM images showed that the morphology of Aβ42 changes from spherical aggregates (oligomers) to fibers in 6 h (Figure 2e,f). No formation of spherical aggregates was observed in the presence of 5 at an equimolar ratio (Figure 2g) and a significantly reduced population of amorphous aggregates was observed after 6 h (Figure 2h). The kinetics of Aβ42 oligomerization was monitored using an ELISA assay in which samples of 2 µM Aβ42 were incubated with 5 at an equimolar ratio for 0, 3, 6, and 12 h and then treated with Aβ oligomer-specific monoclonal antibody (OMAB) (Figure 2i)62. An increase in the absorbance for Aβ42 samples as a function of time was observed, which reflects an increase in the amount of Aβ oligomers (Figure 2i). The absorbance decreased significantly after 6 h suggesting an almost complete conversion of Aβ42 oligomers into the fibrils after 12 h. No significant change in the absorbance was observed for the samples containing Aβ42-5 complex at an equimolar ratio (Figure 2i). As a control, 5 only demonstrated weak binding towards OMAB reflected by weak absorbance (Figure 2i). To validate the results from the ELISA assays26, we performed dot blot analysis of Aβ42 incubated for various durations, in the absence and presence of equimolar 5, and treated with an Aβ42 specific antibody, A11 (Figure 2j). The signal intensity of Aβ42 samples gradually increased from 0 to 6 h and then diminished significantly at 12 h because of the formation of Aβ42 fibrils (Figure 2j). In the presence of 5, no formation of Aβ42 oligomers was observed as reflected by in the weak intensity of the dots throughout the time course of the measurement (Figure 2j). The results from these immunoassays strongly corroborate the TEM data and suggest that 5 effectively suppresses the formation of potential neurotoxic oligomers of Aβ42. Fluorescence titration between 5 and Nα-fluorescein Aβ40 (AβF) was conducted to quantify the binding interaction55. The fluorescence intensity of AβF decreased with increasing amounts of 5, before reaching a plateau (Figure 2k, inset) and a plot of fluorescence intensity as a function of the concentration of 5 yielded a sigmoidal curve (Figure 2k) from which an apparent binding affinity (app. Kd) of 0.44±0.06 µM was calculated reflecting a moderate to strong binding interaction between 5 and Aβ. 6


Page 7 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Kumar et al

Foldamer-induced Structural Arrangement in Aβ. The far UV-CD spectrum of a sample of 20 µM Aβ40 transitioned from that of a random coil to that of a β-sheet conformation in 24 h (Figure 3a). In the presence of 5 at an equimolar ratio, Aβ40 adopts an α-helical conformation, characterized by two minima at ~208 nm and ~222 nm (Figure 3b), which remained constant in the CD even after 24 h (Figure 3b). We observed a positive CD band at ~390 nm, which is the absorbance region for the quinoline chromophore (Figure 3c)63, no such CD band was observed in the sample containing only 5 (Figure 3c). This observation suggests that the interaction with Aβ40 induces an α-helical structure in 5, reflecting a preferential binding to the right handed α-helical conformer (reflected by the positive band) of 5 (Figure 3c)63. Similarly, 5 induces an α-helical conformation in Aβ42 and inhibits the formation of β-sheet conformation (Figure S3). We used HSQC NMR to further investigate the binding between 5 and Aβ. The stoichiometric ratio was restricted to 2:1 (5:Aβ40) to avoid the precipitation at higher concentrations. Using previously published assignments for the chemical shifts of Aβ40 we were able to observe that 5 induces strong secondary structure from residues 13 to 24 without having any effect on the C-terminal residues (Figure 3d,f)64. Aβ40 has been shown to undergo a transition from a weakly folded state to an α-helix conformation in residues 13-24 and 28-36 in the presence of small molecules6 or micelles65. We observed large changes in the chemical shift of residues 13-24, especially His13-Lys16 and Leu17-Phe20, suggesting that this region is a potential binding site of 5. Likely the two negatively charged carboxylate groups of 5 form salt bridges with the positively charged domain of Aβ40 (His13-Lys16), and the two 2-methyl butyl groups interact with and stabilize its hydrophobic domain (Leu17-Phe20). The signals of the Ala21 and Glu22 residues completely disappeared upon addition of 5, indicating that either the amino acid residues are exposed to solvent as a result of conformational readjustment in Aβ40, or fast hydrogen exchange between the protein backbone amide protons of the Aβ40-5 complex and free Aβ40 is equal or faster than NMR time scale measurements. Both NMR and CD clearly suggest a strong conformational change in Aβ40 from a partially folded structure to one with a single helix spanning residues 13 to 24. The critical residues that facilitate Aβ aggregation and neurotoxicity are Phe19/Phe20 and Lys16, since the former is involved in π-π interactions and the latter in both hydrophobic and electrostatic interactions32,

66

. NMR data clearly show that 5

7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 8 of 33

Kumar et al

interacts directly with or in close vicinity to, Lys16 and Phe19/Phe20, induces α-helical conformation, and directs Aβ into amorphous aggregates. 5 Effectively Rescues Aβ-mediated Cytotoxicity The self-assembly and oligomerization of Aβ are associated with the pathology of AD. Since 5 effectively inhibits both these processes, we investigated the cytotoxicity profile of 5 induced structures in Aβ. The cell-based experiments were conducted using mouse neuroblastoma cells (N2a), and cell viability was quantified using CellTiter 96 Aqueous One Solution (MTS) assay. Treatment with 5 µΜ Aβ42 reduced viability of N2a cells to 45% that of controls (Figure 4a). The rank order of inhibition of Aβ-mediated cytotoxicity by tetraquinolines was 5>4≥3>>2≥1 at an equimolar ratio (Figure 4a). The increase in efficacy appears directly related to an increase in the degree of hydrophobicity of the oligoquinolines. At an equimolar ratio, the cell viability in the presence of 5 increased from 45% to 89% (Figure 4a). 5 rescues cytotoxicity in a dose dependent manner with an IC50 of 3.02±08 µM (Figure 4b,c). Oligoquinolines alone were not toxic to N2a cells (Figure S4). The antagonist activity of small molecules against Aβ42 mediated fibrillation and cytotoxicity correlates well, with the exception of EGCG (Figure 4a). In the case of EGCG, which weakly delays Aβ42 fibrillation, potent inhibition of Aβ42 induced cytotoxicity is observed (Figure 4a). We speculate that the mode of action of EGCG and the oligoquinolines is different. EGCG is a known inhibitor of Aβ42 mediated processes; however, its mode of action is partly associated with the remodeling of Aβ42 fibers into amorphous aggregates that are non-toxic in nature27. Oligoquinolines on the other hand, force the monomeric Aβ into an α-helical conformation and modify Aβ mediated processes. Confocal imaging was used to probe the cellular uptake and intracellular localization of 5, Aβ, and the 5-Aβ complex. 5 and Aβ42 were N-terminally labeled with fluorescein (5F) and TexasRed (AβTR), respectively. N2a cells were treated with 5 µM of 5 or Aβ42 doped with 1 µM 5F or AβTR, respectively. Both 5 and Aβ42 were readily taken up by the cells (Figure 5a,b) likely through a recently postulated energy-independent process1. Following cellular uptake, 5 localizes to the mitochondria (Figure 5b) as does Aβ42 (Figure 5a), consistent with reports that Aβ exerts its neurotoxic effects through disruption of mitochondrial function via direct interaction67,68. Under these conditions, cell viability in the presence of 5 µM Aβ42 was ~65% which was 8


Page 9 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Kumar et al

completely rescued (~100%) in the presence of 5 (data not shown). To investigate the mode of action of 5 in rescuing Aβ-mediated cytotoxicity, a 5 µM solution of Aβ42 (20% AβTR) was premixed with 5 µM of 5 (20% 5F) and introduced to the cells. The Aβ42-5 complex was readily taken up by the cells which suggests that 5 does not significantly alter the cell permeable properties of Aβ42 (Figure 5c). To ascertain whether the complex localizes to the mitochondria, a separate experiment was done in which 5 µM Aβ42 (20% AβTR) was premixed with 5 µM unlabeled 5 and introduced to the cells (Figure S5a). As observed with the peptide and compound alone, the labeled Aβ42-unlabled 5 complex colocalizes with mitochondria. Likewise, premixing of 5 µM unlabeled Aβ42 with 5 µM of 5 (20% 5F) yielded a complex that colocalizes with mitochondria (Figure S5b). Taken together, our results suggest that 5 abrogates Aβmediated cytotoxicity by inhibiting formation of oligomeric species that may exert their toxic effects via disruption of mitochondrial function. 5 Eliminates the Preformed Oligomers and Inhibits Prion-like Propagation of Preformed Aβ Seeds. We investigated the effect of 5 on preformed oligomers of Aβ. 5 was added at different time points after the initiation of 5 µΜ Aβ42 amyloid kinetics (Figure 6a). 5 was equally effective in inhibiting Aβ fibrillation when added at 1.5 h or 3 h (during the elongation phase) after the initiation of Aβ42 amyloid kinetics (Figure 6a), but had no effect on the preformed fibers (Figure 6a). TEM image analysis show that 5 directs Aβ42 oligomers to off pathway structures when added 1 h and 2 h after the start of Aβ42 amyloid kinetics (Figure 6b,e). ELISA and dot blot assays confirm the formation of 2 µΜ Aβ42 oligomers after 3 h and 6 h with gradual increase in absorbance and chemiluminescence intensity, respectively (Figure 6f,g). In marked contrast, no evidence of oligomers was observed when 5 was added 3 h and 6 h after the start of Aβ42 amyloid reaction at an equimolar ratio (Figure 6f,g). The effect of 5 was also tested on Aβ42 oligomer induced toxicity in N2a cells. Solutions of 5 µΜ Aβ42 were incubated in buffer for 0, 1, 2, and 3 h and then introduced to the cells, which reduced cell viability to 45%, 41%, 55%, and 54%, respectively (Figure 6h). In the presence of 5, the cell viability for these samples increased to 87%, 84%, 93%, and 93%, respectively (Figure 6h),suggesting that 5 has the tendency to disrupt the preformed oligomer assembly. 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 33


Kumar et al

Preformed fibers (seeds) of Aβ accelerate the fibrillation via secondary nucleation mechanisms69, 70

. The secondary nucleation process generates the key neurotoxic oligomers and therefore

inhibition of these processes is essential from a therapeutic point of view69,

70

. Addition of

preformed fibers (10%, v,v) to freshly aliquoted 5 µM Aβ42 accelerates the fibrillation by decreasing the t50 from 3.38±0.33 h to 0.35±0.08 h (Fig 6i; Figure S6). In the presence of 5, the seed catalyzed fibrillation of Aβ42 was completely suppressed (Figure 6i,j). Similarly, the aggregation for de novo and seed-catalyzed (25%, v,v) Aβ40 yielded t50 values of 42.1±4.8 h and 14.8±3.1 h, respectively (Figure S7), whereas the seed catalyzed aggregation of Aβ40 was completely suppressed in the presence of 5 (Figure S7). TEM image of seed-catalyzed Aβ42 aggregation displayed an abundance of fibers after 12 h (Figure 6k). In contrast, only amorphous aggregates were observed in the presence of 5 after 12 h (Figure 6l). The effect of 5 on seedinduced secondary toxicity of Aβ42 was also assessed. Cell viability decreased from 45% to 30% in the presence of 5 µΜ Aβ42 and 5 µΜ Aβ42+seeds (10%, v,v) respectively (Figure 6j). Remarkably, the cell viability recovered from 30% to 81% in the presence of 5 (Figure 6j). We also investigated the seeding properties of the amorphous aggregates of Aβ generated in the presence of 5. Sample of the Aβ42-5 complex was centrifuged (filter size = 3 kDa) to remove the unbound 5 from the Aβ42-5 complex. While, the addition of fibers (10%, v,v) to monomeric 5 µM Aβ42 accelerate fibrillation by ~8 fold, the amorphous aggregates (generated in the presence of 5) inhibit Aβ aggregation for more than 10 h (Figure 6 m,n) as further confirmed by TEM (Figure S8). Other studies have reported similar inhibitory behavior of amorphous aggregates generated by other conditions such as salt concentration71 or ultra-sonication72. Our results suggest that 5 generates stable amorphous aggregates which are seed incompetent structures that function as a sink for the monomeric Aβ and inhibit its fibrillation.

10


Page 11 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Kumar et al

Conclusions The promiscuity of Aβ conformations, derived from both monomer and the oligomer, make it a daunting task to develop Aβ structure specific ligands. Strategies that could potentially alter the structure of the monomeric or smaller oligomeric species, thereby preventing formation of toxic oligomers, are considered viable approaches from a therapeutic point of view. Such strategies have been employed by us and others, whereby small molecules/ proteins were designed to induce a secondary structure in Aβ while simultaneously adapting to the conformational preferences of Aβ to achieve optimal surface contacts and, in doing so, modulating Aβ-mediated downstream toxic functions. Here, we combine both of these properties in an oligoquinoline, 5, which interacts with Aβ in a sequence and structure specific manner and affects various facets of Aβ aggregation across both biophysical and cellular assays. Presumably, the carboxylate groups form a salt bridge with Lys16 and the higher order of specificity is achieved by the interaction with His13, His14 and the hydrophobic domain (Leu17-Phe20), thereby inhibiting the interactions that facilitate Aβ oligomerization. The Aβ structures stabilized/induced by 5 are non-toxic and do not exhibit prion-like properties, which underlines the potential of the oligoquinoline as a ligand for the therapeutic intervention of AD. This study is novel compared to previously published inhibitors of Aβ in many respects: (1) compound 5 adjust its conformation to optimize interaction with Aβ; (2) a clear binding site is detected in the presence of compound 5 on Aβ, which could aid in developing more effective inhibitors; (3) the surface properties of the oligoquinolines can conveniently be modified with a more versatile chemical variability than proteins, using synthetic approaches, without perturbing their folded conformation; (4) compound 5 not only manipulates the monomeric conformation of Aβ, it also disrupts the preformed neurotoxic Aβ oligomers and represents an interesting therapeutic approach; (5) using confocal microscopy, we were able to demonstrate the mode of action of compound 5 on Aβ mediated fibrillation and cytotoxicity; (6) using confocal image analysis, the current study also clearly demonstrates that compound 5 retains its antagonist activity towards Aβ mediated toxicity in cellular milieu, which is unprecedented for Aβ inhibitors73. The conveniently achievable protein surface mimicking properties, allied with high target specificity, and tendency to adopt to a target specific conformations, make oligoquinolines an important class of therapeutic molecules. 11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 33


Kumar et al

Materials and Methods Materials. All peptides used in the study (Aβ40, Aβ42) were purchased with >98% purity from Anaspec (Fremont, CA, USA) and used without further purification.

15

N- labeled wild-type

human Aβ40 was purchased from rPeptide (Bogart, GA, USA) and used without further purification. ThT was purchased from Acros Organics (NJ, USA). The 96-well plates (black, flat bottom) were purchased from Corning Coster (Corning, NY, USA). All chemicals were purchased from commercial suppliers and used without further purification. Silica plates (with UV254, aluminum backed, 200 µm) and silica gel (standard grade, particle size = 40-63 micron, 230 × 400 mesh) for flash column chromatography were purchased from Sorbent Technologies (Norcross, GA, USA). Dry solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA). Alkyl iodides, alkyl alcohols, anhydrous triethylamine, 2-chloro-1-methylpyridinium iodide, tertbutyl bromoacetate, HPLC grade trifluoroacetic acid, and triethylsilane (TES) were purchased from Sigma Aldrich (St. Louis, MO, USA). 5-(Aminoacetamido) Fluorescein was purchased from ThermoFisher Scientific (Waltham, MA) and used without further purification. Peptide Preparation. Aβ40 and Aβ42 were dissolved in 1,1,1,3,3,3-hexafluoroisopropanol (HFIP) (1 mg/mL) and kept at r.t. for 1 h with occasional vortexing to maintain them in monomeric state. Peptides were then aliquoted into small fractions and lyophilized overnight and stored at -80 °C until use. Peptide concentration was checked by dissolving one of the aliquots in water and measuring absorbance at 280 nm. For solution-based assays, lyophilized samples were allowed to equilibrate at r.t. for 20 min. and then dissolved in pure DMSO (Amresco, Solon, OH, USA). The concentration of the stock solution for all the biophysical assays was 0.5-1.0 mM. Synthesis of the Oligoquinolines. The synthetic protocol and the characterization of the relevant compounds are presented in detail in the supplementary information. Thioflavin T (ThT)-based Aggregation Kinetic Assay. Kinetic assays were conducted on a FlexStation 3 Multi-Mode Microplate reader from Molecular Devices (Sunnyvale, CA, USA). Experiments were conducted in triplicate in a 96-well plate with a final volume of 200 µL per well. Every measurement was an average of 50 readings. The aggregation of amyloidogenic peptide (Aβ40 and Aβ42) was initiated by addition of peptide from a stock solution (in DMSO, 0.5-1 mM) to phosphate buffer. The final concentration of each peptide was different based on 12


Page 13 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Kumar et al

their aggregation (see main text). The stoichiometry of ThT dye for each peptide was 0.5:1 (ThT:peptide). Peptide aggregation was monitored by ThT fluorescence (λex = 445 nm and λem = 485 nm). The blank sample contained everything except peptide. The sample data were processed by subtracting the blank and renormalizing the fluorescence intensity by setting the maximum value to one. Kinetic assays in the presence of small molecules were conducted under the same conditions except that the small molecules were added from a stock solution (1 mM or 10 mM in DMSO) at an equimolar ratio and the final concentration of DMSO was less than 1.0% (v/v). Small molecules were added to the wells with ThT and buffer and mixed gently with a pipette before adding the peptide. To keep the conditions identical, an equal amount of DMSO was added to the wells with the peptide only reactions. Kinetic profiles were processed using Origin (version 9.1). Kinetic curves were fit using the built-in sigmoidal fit. Each run was fit independently to extract the t50 (time required to reach 50% of the maximum fluorescence intensity). Error bars represent standard deviations from the mean of at least three independent experiments. Seed-Catalyzed Kinetic Assay. Seeds of Aβ40/Aβ42 were prepared by incubating 200 µM of Aβ40/Aβ42 in phosphate buffer at r.t. The samples were aged for 48 h and the formation of fibers was confirmed by TEM and ThT before storage at -20 °C until use. For the seed catalyzed aggregation of Aβ40, 25% (based on the monomeric Aβ40, v/v) seeds were added along with ThT to phosphate buffer to the 96-well plate. The aggregation was initiated by the addition of monomeric 20 µM Aβ40 followed by gentle mixing. The process was similar in the case of Aβ42 except that the seed concentration was 10% (based on the monomeric Aβ42, v/v) and the monomer Aβ42 was 5 µM. The seed catalyzed TEM image analysis was performed the same way as described for ThT kinetic assay except no ThT dye was added in the solution. The seeds were generated by aging 200 µM Aβ40/Aβ42 for 48 h in phosphate buffer with occasional vortexing. The concentration of seeds was calculated based on the monomeric concentration of Aβ40/Aβ42. Error bars represent standard deviations from the mean of at least three independent experiments. Transmission Electron Microscopy (TEM) Analysis. Aβ42 (5 µM) was incubated in phosphate buffer in the absence and presence of 5 at various time intervals at an equimolar ratio. Aliquots of these samples were then applied to glow-discharged carbon-coated 300-mesh copper grids for 13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 14 of 33

Kumar et al

2 min and dried. Grids were negatively stained with uranyl acetate (2%, w/v) and dried. Micrographs of grids were examined on a Phillips CM12 Cryoelectron Microscope equipped with Gatan 4k × 2.7k CCD camera at 120-kV accelerating voltage. Ultraviolet-Visible (UV-Vis) Spectroscopy. UV-Vis was employed to probe the chemical aggregation of 5 under kinetic assay conditions. The measurements were carried out on a double beam Carry 100 Bio spectrophotometer (Agilent Technologies, Santa Clara, CA) controlled by Cary WinUV software (version = 3.0). A stock solution of 5 (10 mM in DMSO) was subjected to a series of dilutions with a concentration range from 10 to 150 µM (Total DMSO = 1.5% (v,v) in 150 mM KCl, 50 mM NaPi, pH 7.4). The spectra of 5 were recorded at 1.0 nm intervals from 500 to 200 nm with a scan rate of 300 nm/min. Subsequently, the solutions of 5 were filtered (size = 0.2 µm, VWR sterile syringe filter) and the spectra were recorded from 500 to 200 nm. Dot Blot Assay. Samples of 2 µM Aβ42 were incubated at various time intervals (see main text for detail) in the absence and presence of 5 at an equimolar ratio. The samples were then applied to a nitrocellulose membrane and dried at r.t. for 1 h or overnight at 4 °C. The membranes were then blocked with 5% nonfat milk in Tris buffer (20 mM Tris, pH 7.4) for 1 h at r.t. The nitrocellulose membranes were then washed (× 3) with 20 mM Tris, pH 7.4 supplemented with 0.01% Tween-20 (TBST) and incubated with polyclonal A11 antibody (1/1000 dilution in 5% nonfat milk in TBST, Life Technologies Corp., Grand Island, NY, USA) overnight at 4 °C. Samples were then washed (× 3) with TBST buffer and incubated with horseradish peroxidase (HRP) conjugated anti-rabbit IgG (1/500 dilution in 5% nonfat free milk in TBST) at r.t. for 1 h. The dot blots were then washed with TBST buffer (× 3), developed using the ECL reagent kit (Amersham, Piscataway, NJ, USA), and imaged using a Typhoon FLA 9000 instrument (GE Healthcare Life Sciences, Pittsburgh, PA, USA) using chemiluminescence settings. A similar experiment was repeated using 6E10 antibody (1/1000 dilution in 5% nonfat free milk in TBST, Biolegend, San Diego, CA, USA) for comparison. ELISA Assay. A Nunc-Immuno MaxiSorp plate (Sigma Aldrich, St. Louis, MO, USA) was incubated with 2 µg mL-1 Aβ oligomer-specific antibody (OMAB, Agrisera, Sweden) in PBS buffer overnight at 4 °C (200 µl/ well). Wells were blocked with 5% fat-free milk in PBS buffer with 0.1% Tween 20 (PBST) for 1 h at 4 °C and washed with PBST buffer (× 3). Wells were 14


Page 15 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Kumar et al

then treated with Aβ42 samples overnight at 4 °C. Aβ42 samples were prepared by incubating 30 µM Aβ42 in the absence and presence of 5 at an equimolar ratio at various time intervals at r.t. Samples were diluted by 1:15 in PBS buffer before adding to a 96-well plate. After adding Aβ42 samples, the wells were washed (× 3) with PBST buffer followed by the addition of 6E10 antibody (1/1000 dilution in 5% nonfat free milk in PBST buffer) for 1 h at r.t. Wells were washed (× 5) with PBST buffer and treated with an anti-mouse HRP-conjugated IgG (1/10,000 dilution in 5% nonfat free milk in PBST buffer). Wells were washed (× 5) with PBST buffer and treated with TMB Peroxidase EIA Substrate Kit (Biorad, Hercules, CA, USA). The plates were developed until the color of the solution turned blue. The reaction was stopped by adding 100 µL of 1N H2SO4 to each well. The color of the solution changed to yellow from blue. The absorbance was recorded at 450 nm on a FlexStation 3 Multi-Mode Microplate reader from Molecular Devices (Sunnyvale, CA, USA). The experiment with each sample was repeated in triplicate. Error bars represent standard deviations from the mean of at least three independent experiments. Circular Dichroism (CD) Spectroscopy. A freshly prepared stock solution of Aβ40 (1 mM) was diluted to 20 µM in phosphate buffer for CD measurements. The spectra of Aβ40 were recorded at 0.5 nm intervals from 190 to 260 nm with an averaging time of 10 sec. and an average of three repeats on a Aviv Stopped Flow CD Spectropolarimeter (Model 202SF). Spectra were recorded from 190 to 600 nm using the identical method as described above, except 20 µM Aβ40 was diluted in the solution of 5 in phosphate buffer at an equimolar ratio. Phosphate buffer: 150 mM KCl, 50 mM NaPi, pH 7.4. All the experimental conditions were same in the case of Aβ42 except the concentration, which was 15 µM. Fluorescence Titration. Nα-amino-terminal fluorescein-labeled Aβ40 (AβF) was purchased from Anaspec (Fremont, CA, USA) and used without further purification. To ensure the monomeric state of AβF, the peptide was treated similarly to other peptides and stored at -80 °C in the dark until use. Fluorescence measurements were performed on a FlexStation 3 Multi-Mode Microplate reader from Molecular Devices (Sunnyvale, CA, USA). Fluorescence titrations were conducted in triplicate in a 96-well plate with a final well volume of 200 µL. For fluorescence measurements, the fluorescein dye was excited at 492 nm and the spectra were recorded from 500 nm to 600 nm. A 10 nM AβF solution in fluorescence assay buffer (20 mM NaPi, 1% TFE, 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 33


Kumar et al

pH 7.4) was titrated with incremental amounts of 5 (in DMSO, 0.5-2 mM) and the spectra were recorded from 500 nm to 600 nm. A number of high concentration stock solutions of 5 were prepared to minimize the amount of DMSO in the fluorescence titrations (

Foldamer-Mediated Structural Rearrangement Attenuates AÎ²

Recommend Documents