Bikash R.
Sahoo
a,
Takuya
Genjo
a,
Takahiro W.
Nakayama
b,
Andrea K.
Stoddard
a,
Toshio
Ando
b,
Kazuma
Yasuhara
c,
Carol A.
Fierke
ad and
Ayyalusamy
Ramamoorthy
*a
aBiophysics and Department of Chemistry, University of Michigan, Ann Arbor, MI 48109-1055, USA. E-mail: ramamoor@umich.edu
bBio-AFM Frontier Research Center, Kanazawa University, Kanazawa 920-1192, Japan
cGraduate School of Materials Science, Nara Institute of Science and Technology, Ikoma, Nara 6300192, Japan
dDepartment of Chemistry, Texas A&M University, College Station, TX 77843, USA
First published on 27th February 2019
In humans, β-amyloid and islet amyloid polypeptide (IAPP, also known as amylin) aggregations are linked to Alzheimer's disease and type-2 diabetes, respectively. There is significant interest in better understanding the aggregation process by using chemical tools. Here, we show the ability of a cationic polymethacrylate-copolymer (PMAQA) to quickly induce a β-hairpin structure and accelerate the formation of amorphous aggregates of β-amyloid-1-40, whereas it constrains the conformational plasticity of amylin for several days and slows down its aggregation at substoichiometric polymer concentrations. NMR experiments and microsecond scale atomistic molecular dynamics simulations reveal that PMAQA interacts with β-amyloid-1-40 residues spanning regions K16-V24 and A30-V40 followed by β-sheet induction. For amylin, it binds strongly close to the amyloid core domain (NFGAIL) and restrains its structural rearrangement. High-speed atomic force microscopy and transmission electron microscopy experiments show that PMAQA blocks the nucleation and fibrillation of amylin, whereas it induces the formation of amorphous aggregates of β-amyloid-1-40. Thus, the reported study provides a valuable approach to develop polymer-based amyloid inhibitors to suppress the formation of toxic intermediates of β-amyloid-1-40 and amylin.
5 μM of β-amyloid-1-40 monomers dissolved in 10 mM sodium phosphate, pH 7.4, or 5 μM of amylin monomers dissolved in 30 mM sodium acetate, pH 5.5, were incubated for 24 hours under continuous agitation with/without an equimolar concentration of PMAQA. β-amyloid-1-40 or amylin fibers were prepared by incubating 5 μM of peptide monomers in respective buffers as described above at room temperature for 3 days under continuous agitation. Then, the fibers were sonicated using an ultrasonic bath sonicator for 2 hours at room temperature to generate short fibers, which were used as amyloid seeds. The sonicated fibers were incubated with 5 μM of PMAQA in respective peptide buffers at room temperature for 24 hours under continuous agitation to check the effect of the polymer on the fiber morphology. In addition, to monitor the effect of PMAQA on the self-seeding reaction, sample solutions containing sonicated fibers were incubated with 5 μM monomers of respective peptides for 24 hours in the presence or absence of 5 μM PMAQA. 10 μL of all sample solutions described above were added to a collodion-coated copper grid and incubated for 3 minutes at room temperature followed by three times rinsing with double deionized water. Then, the copper grid was stained with 3 μL of 2% (w/v) uranyl acetate and incubated for 2 minutes followed by three times rinsing with double deionized water. The sample grids were dried overnight under vacuum and used to obtain transmission electron microscopy (TEM) images using a HITACHI H-7650 transmission microscope (Hitachi, Tokyo, Japan) at 25 °C.
Fig. 1 Effect of PMAQA on the aggregation kinetics of β-amyloid-1-40 or amylin. Relative ThT fluorescence of 5 μM β-amyloid-1-40 in 10 mM sodium phosphate, pH 7.4, (a) or amylin in 30 mM sodium acetate, pH 5.5, (b) in the presence of variable PMAQA concentrations at the indicated peptide to polymer molar ratios. The data in both graphs represent the average of ThT values performed in triplicate as shown in Fig. S2.† |
Unlike the effect of PMAQA on β-amyloid-1-40 fibrillation, amylin aggregation was slowed down at substoichiometric PMAQA concentrations, i.e. an amylin:PMAQA molar ratio of 1:0.02 or 1:0.04 (Fig. S2c†). Increasing the polymer concentration to a 1:0.2 amylin:PMAQA molar ratio further slowed down the aggregation with a lag-time of over 2 days as compared to its aggregation in ∼1 day in the absence of PMAQA (Fig. 1b). A further increase in PMAQA concentration, i.e. at equimolar polymer:peptide or superstoichiometric PMAQA concentrations, amylin aggregation was significantly slowed down as revealed by the negligibly small ThT fluorescence intensity (Fig. 1b and S2d†). The observed increase in the ThT intensity in both peptides at substoichiometric PMAQA concentrations (peptide:polymer 1:0.5 or lower) indicates the presence of a large number of β-amyloid or amylin fibers. On the other hand, at superstoichiometric PMAQA concentrations (peptide:polymer 1:1 or higher), a substantial decrease in the ThT intensity indicates the presence of a smaller number of fibers. Specifically, the low ThT intensity observed for β-amyloid-1-40 could also indicate the formation of morphologically distinct fibers, i.e. amorphous like aggregates that depict low ThT fluorescence intensity.51
Next, the conformational changes in both peptides in the presence and absence of PMAQA were investigated using far-UV CD experiments. The CD spectra revealed a gradual structural transition (from unfolded to folded) in β-amyloid-1-40 when titrated with a 1:0.01 or 1:0.04 β-amyloid:PMAQA molar ratio as indicated by a reduction in CD molar ellipticity [Θ] and a small change in CD minima at ≈200 nm (Fig. 2a). The reduction in [Θ] indicates β-amyloid-1-40 aggregation and is in line with the observed ThT aggregation (Fig. 1a). A partial helical CD spectrum containing α/β secondary contents of 9.4/26.1% (as estimated by BESTSEL52) was observed at a 1:0.2 β-amyloid:PMAQA molar ratio (Fig. 2a). A further increase in PMAQA concentration to a 1:1 β-amyloid:PMAQA molar ratio marginally increased the β-sheet contents in β-amyloid-1-40 (α/β content of 7.7/28.2%). The shift of CD minima at ≈200 nm and decrease in [Θ] with an increasing concentration of PMAQA indicated its activity in accelerating β-amyloid-1-40 aggregation as observed in the ThT assay (Fig. 1a). Remarkably, unlike β-amyloid-1-40, no significant conformational changes were observed in amylin when titrated with 0.2 to 2 equivalents of PMAQA with a CD minimum centered at ≈200 nm with a very little change in [Θ] that represents a partially folded random-coil rich conformation (Fig. 2b).
We next monitored the stability of the PMAQA induced β-amyloid-1-40 or amylin secondary structures for 5 days by incubating the peptides at room temperature in the absence or presence of 1.5 molar equivalents of PMAQA (Fig. 2c–f). The time-lapse CD spectra of β-amyloid-1-40 or amylin in the absence of PMAQA showed a sequential structural transition from a random-coil rich (negative peak ≈200 nm) to β-sheet structure (Fig. 2c and d). While β-amyloid-1-40 in solution is reported to exhibit a sequential structural change to form cross β-sheet structures over several days (Fig. 2c),53 in the presence of 1.5 molar equivalents of PMAQA a rapid β-sheet induction within several minutes was observed (Fig. 2e). The observed rapid structural change correlates with the ThT aggregation kinetics (Fig. 1a) and indicates a reduction in the level of potential β-amyloid-1-40 oligomers that are reported to be neurotoxic.53,54
Similar to the amyloidogenic properties of β-amyloid-1-40, amylin in solution after 2 days depicted a CD spectrum with a maximum at ≈200 and a minimum at ≈218 nm indicating its transitory state characterized by an increasing percentage of β-sheets (54%) (Fig. 2d). But, unlike β-amyloid-1-40, and as observed in ThT assays for amylin (Fig. 1b), CD spectra showed a relatively slow change in the secondary structure of amylin when incubated with a 1.5 molar excess of PMAQA. The CD minimum was observed at ≈200 nm up to day 3 and the spectra were quantified using BESTSEL (Fig. 2f and Table 1). Secondary structure assessment of PMAQA bound amylin from CD spectra showed a relative increase in α-helix (14.5%) and decrease in β-sheet (21.7%) content as compared to that observed in the absence of PMAQA (Table 1). In addition, an increased percentage of parallel β-sheets (13.4% on day 5) was observed for amylin in the presence of PMAQA over time as compared to 0% (on day 5) in the absence of PMAQA. This observation indicates that PMAQA bound amylin could have both anti-parallel and parallel β-structures. Such observations have been found previously using X-ray crystallography in different segments of amylin (segments 13–18 with parallel and anti-parallel β-structures for segments 16–21, 22–28, and 23–29).55 On the other hand, β-amyloid-1-40 showing a positive CD band at ≈200 nm and a negative CD band at ≈225 nm in the presence of PMAQA indicates the formation of a predominant β-sheet structure (Fig. 2e). The observed CD minimum with a red-shift (from ≈222 on day 1 to ≈225 nm on day 5) in β-amyloid-1-40 correlates with previously observed β-sheet rich supramolecular structures in modified β-amyloid-1-40 and other small peptide aggregates.56,57
Time | 0 d | 1 d | 2 d | 3 d | 4 d | 5 d |
---|---|---|---|---|---|---|
a Antiparallel β-sheet. b Parallel β-sheet. | ||||||
Amylin in 30 mM sodium acetate, pH 5.5 | ||||||
α-helix | 0.7 | 1.4 | 0 | 0 | 1.3 | 0.4 |
βa | 42.1 | 37.5 | 54.0 | 59.0 | 53.1 | 53.0 |
βb | 0 | 3.9 | 0 | 0 | 0.2 | 0 |
Amylin+PMAQA in 30 mM sodium acetate, pH 5.5 | ||||||
α-helix | 14.5 | 9.1 | 10.6 | 5.3 | 5.3 | 6.6 |
βa | 21.7 | 29.7 | 28.1 | 30.3 | 33.4 | 25.4 |
βb | 0 | 0 | 0 | 5.7 | 3.7 | 13.4 |
Next, we carried out FT-IR experiments on samples after 6 hours of incubation under similar experimental conditions used for CD measurements to further confirm the observations from CD experiments. As shown by the FT-IR spectra in Fig. 2g, predominant β-sheet structures of β-amyloid-1-40 were observed as indicated by a sharp amide I peak at 1628 cm−1 along with an increasing percentage of turns (at 1675 cm−1) in the presence of PMAQA, which are in qualitative agreement with CD results shown in Fig. 2e. In the case of amylin, FT-IR spectra showed a minor change in the characteristics of the helical band at 1650 cm−1 in the presence of PMAQA (Fig. 2h). This indicates that PMAQA binding restrains the conformation of partially folded amylin.43 The β-sheet rich β-amyloid and partially folded amylin conformations in the presence of PMAQA (Fig. 2) coupled with ThT observations (Fig. 1) indicated the presence of distinct peptide species such as fibers or low-order aggregates. To probe this, SEC profiling was performed to quantify the size distribution of the PMAQA bound peptide species. SEC analysis of β-amyloid-1-40 incubated with a 1.5 molar excess of PMAQA for ∼5 minutes at room temperature showed two different elution profiles. The fractions collected at ∼5 to 12 mL correspond to amyloid fibers or protofibers and those collected at ∼20 to 25 mL are free polymers or low-order or monomeric β-amyloid-1-40 (Fig. S3a†).58 Remarkably, the SEC profile for PMAQA-amylin mixed solution, incubated for a relatively longer time (overnight) at room temperature, exhibited low-order or monomeric amylin and free polymers eluted at ≈20 to 25 mL (Fig. S3b†).59 Taken together, the above described experimental results present a counteractive role of PMAQA in β-amyloid-1-40 and amylin aggregation.
Next, we studied the binding mechanism of PMAQA with β-amyloid-1-40 or amylin using an integrated NMR and MD simulation approach. The amide-NH region of 1H NMR spectra obtained in the absence of PMAQA showed monomer or low-order aggregates for both β-amyloid-1-40 and amylin as indicated by the dispersed NMR peaks (Fig. 3). A substantial change in the 1H NMR spectra of β-amyloid-1-40 was observed at a 1:0.02 β-amyloid:PMAQA molar ratio (Fig. 3, pink trace). Remarkably, the loss of amide-1H peaks of β-amyloid-1-40 was observed when the PMAQA concentration (1:0.05 β-amyloid:PMAQA ratio) was slightly increased (Fig. 4d). This agrees with the PMAQA induced β-amyloid-1-40 aggregation observed in ThT assays. In contrast, the amide peaks of amylin were observed even when titrated with a 1.2 molar equivalent of PMAQA (Fig. 3, blue trace). These NMR findings are in good agreement with the observed conformational transition from CD and ThT based aggregation results (Fig. 1b and 2b). Interestingly, a significant line broadening was observed for the proton peak of the –NR3+ group of PMAQA (at 2.97 ppm) in the presence of β-amyloid-1-40 which indicates the interaction of –NR3+ with β-amyloid-1-40. In contrast, a sharp proton peak for –NR3+ of PMAQA was observed in amylin solution (Fig. 3). This observation most likely indicates the formation of an electrostatic interaction between the cationic PMAQA and anionic β-amyloid-1-40, as β-amyloid-1-40 (at pH = 7.4) and amylin (at pH = 5.5) carry negative (−3) and positive (+4) charges, respectively.60,61 Thus, while the cationic group of PMAQA binds strongly to anionic β-amyloid-1-40, a repulsive force could be expected in the presence of cationic amylin.
To gain further atomistic insight into the mechanism, β-amyloid-1-40 aggregation was monitored using a 2D 15N/1H SOFAST-HMQC experiment at variable substoichiometric polymer concentrations; since 1H NMR of amylin showed very little change in the amide region at superstoichiometric PMAQA concentrations, further NMR experiments were not carried out on amylin. As shown in Fig. 4a–d, intensity reduction and chemical shift changes were observed for β-amyloid-1-40 residues indicating a PMAQA induced structural rearrangement for β-amyloid-1-40. A substantial loss of signal intensity was observed in β-amyloid-1-40 spanning regions K16-V24 and A30-V40 (as per previously published assignments62) indicates potential PMAQA binding sites on β-amyloid-1-40 at 1:0.01 and 1:0.02 β-amyloid:PMAQA molar ratios (Fig. 4b and c). A further increase in PMAQA concentration (1:0.05 β-amyloid:PMAQA molar ratio) increased the line-broadening and also resulted in the loss of 15N/1H resonances (Fig. 4d). This indicates that PMAQA interaction with β-amyloid-1-40 induces the aggregation of the peptide.
To further explore the binding mechanism of PMAQA with amylin or β-amyloid-1-40 at the structural level, we performed all-atom MD simulation on a time scale of 0.7 or 1 μs, respectively. Structural analysis showed the formation of a substantial number of hydrogen bonds between PMAQA and β-amyloid-1-40 over a time-scale of 1 μs (Fig. S4a†). All-atom MD simulation revealed potential electrostatic interactions or hydrogen bonds between the Glu22 and Asp23 residues of β-amyloid-1-40 and –NR3+ of the polymer (Fig. 5a), whereas no interaction with the charged residues of β-amyloid-1-40 in the N-terminal (Asp1, Glu3 and Asp7) was observed. Other residues such as Leu17, Phe20, Val24, Ser26, Asn27, Ile31, and Ile32 were observed to interact with PMAQA through hydrogen bonding or hydrophobic interactions (Fig. 5a and Table S1†). This correlates with the NMR results that showed a substantial decrease in the NMR signal intensity in β-amyloid-1-40 regions spanning K16-V24 and A30-V40 (Fig. 5c). MD calculations also revealed several intermolecular hydrogen bonds and hydrophobic interactions between amylin and PMAQA (Fig. S4a†). Amylin residues such as Asn21, Ile26, Ser28, Thr30, Asn31, Ser34 and Tyr37 were identified to be involved in intermolecular hydrogen bonding interactions with PMAQA (Fig. 5b and Table S2†). Overall, PMAQA exhibited a greater number of hydrogen bonds with amylin than with β-amyloid-1-40 indicating a relatively stronger binding affinity of PMAQA to amylin (Fig. S4a†).
Fig. 5 Structural insights into the PMAQA interaction with β-amyloid-1-40 or amylin. MD snapshots showing PMAQA (shown in balls and sticks) interaction with β-amyloid-1-40 (a) or amylin (b) as a cartoon. The PMAQA binding peptide amino acids (shown in sticks) are labelled and hydrogen bonds are shown in black dashed lines in Discovery Studio Visualizer. (c) Peak intensities measured from the 2D SOFAST-HMQC spectra (Fig. 4) of 60 μM β-amyloid-1-40 in the absence or presence of 0.6 μM PMAQA. The yellow area highlights the β-amyloid-1-40 residues with significantly reduced signal intensities. (d) Root mean square fluctuation (RMSF) of residues in β-amyloid-1-40 (grey) or amylin (red) interacting with PMAQA derived from 1 or 0.7 μs MD simulations, respectively. The yellow region indicates a comparatively flexible β-amyloid-1-40 N-terminal domain. The blue and orange regions indicate aggregation core domains of amylin and β-amyloid-1-40, respectively. The average RMSF value is shown using dashed horizontal lines. |
Computation of binding energy using MM/PBSA presented distinct energetic parameters that favor PMAQA binding to β-amyloid-1-40 or amylin. The binding energy estimated for PMAQA-β-amyloid and PMAQA-amylin complexes was −16.5 ± 10.1 and −27.4 ± 11.7 kcal mol−1, respectively. The energetic parameter analysis shown in Table 2 indicates that coulombic and van der Waals interactions play a crucial role in PMAQA binding to β-amyloid-1-40, while the polar solvation energy restricts the interaction. On the other hand, PMAQA binding to amylin is favored by the nonpolar component of the solvation free energy and van der Waals interaction, while coulombic interaction disfavors the complex formation (Table 2). This indicates that the opposite charges in β-amyloid-1-40 and amylin are the major factors responsible for their distinguished amyloid aggregation behavior in the presence of PMAQA. The electrostatic interaction or repulsion in β-amyloid-1-40 or amylin interaction with PMAQA, respectively, could contribute to the acceleration or deceleration of amyloid aggregation as observed in previous studies using differently charged nanoparticles.63,64
Polar contribution | Non-polar contribution | ||||||
---|---|---|---|---|---|---|---|
ΔGbinda | ΔGcoulb | ΔGpsc | ΔGpod | ΔGvdwe | ΔGnpsf | ΔGnpg | |
a Binding free energy. b Coulombic term. c Polar solvation. d Polar contribution. e van der Waals energy. f Nonpolar solvation. g Nonpolar contribution. | |||||||
β-amyloid + PMAQA | −16.5 ± 10.1 | −521.4 ± 132.7 | 560.0 ± 137.0 | 38.5 | −48.0 ± 7.2 | −7.0 ± 0.8 | −55.1 |
Amylin + PMAQA | −27.4 ± 11.7 | 84.9 ± 106.2 | −50.0 ± 103.6 | 34.8 | −54.7 ± 7.8 | −7.6 ± 0.9 | −62.3 |
The root mean square deviation (RMSD) of the backbone atoms calculated from 1 μs MD simulation of the β-amyloid-PMAQA system showed an RMSD plateau with an average value ≈11 Å (Fig. S4b†). A relatively small backbone RMSD (average value ≈8.8 Å) was observed for the amylin-PMAQA system indicating a comparatively stable complex formation (Fig. S4c†). The RMSD of PMAQA calculated from all-atom MD simulation depicted a nearly equal RMSD value of ≈6.5 Å for both MD systems (Fig. S4d†). Overall, the stable peptide backbone with an average RMSD of <2 Å over several hundreds of nanoseconds indicates a strong coupling between PMAQA and amylin or β-amyloid-1-40 over the ∼μs time scale MD simulation. Furthermore, the root mean square fluctuation (RMSF) analysis of individual amino acids in β-amyloid-1-40 or amylin highlighted the potential PMAQA interaction regions. The amyloid aggregation core domains in both β-amyloid-1-40 (KLVFF) and amylin (NFGAIL) depicted an RMSF value lower than their corresponding average values as indicated in Fig. 5d. In amylin, residues 7–16 showed the lowest RMSF values that folded to a stable α-helical conformation. This indicates that the PMAQA interaction restrains the structural and dynamic properties of amylin. The tight coupling of PMAQA close to the proximity of the NFGAIL region in amylin followed by restriction in protein structural rearrangement reveals mechanistic insights into the antagonist properties of PMAQA in amylin aggregation. Furthermore, the secondary structure analysis performed using MD trajectories in the amylin-PMAQA complex showed no significant secondary structural change during the MD simulation in amylin (Fig. S5,† bottom). On the other hand, a substantial secondary structural change was observed in β-amyloid-1-40 complexed with PMAQA including the induction of an antiparallel β-sheet along the terminal residues (Fig. 5a and S5,† top). The interaction of PMAQA with the centrally located residues of β-amyloid-1-40 such as Glu22, Asp23, Val24, Ser26 and Asn27 induced a random-coil and helix conformation in β-amyloid-1-40. Structural changes in β-amyloid-1-40 by affecting these central residues by mutation, disruption of the D23-K28 salt-bridge or binding of the hexapeptide of genetic β-amyloid-1-40 variants has been reported previously to be crucial in accelerating β-amyloid-1-40 aggregation and induction of the terminal β-structure.65–68 The charge–charge interaction between PMAQA (–NR3+) and Glu22/Asp23 in β-amyloid-1-40 could interfere with the early oligomer morphology and distribution of large β-amyloid-1-40 aggregates as observed in SEC (Fig. S3†).69 The binding of PMAQA to Glu22 and Val24 also affects the hydrophobic interaction between Val24 and Lys28 and electrostatic interaction between Lys28 and Glu22/Asp23.70
We next monitored the aggregation kinetics of amylin in real time using HS-AFM as the results presented above indicated an antagonistic behavior of PMAQA in amylin aggregation (Fig. 6). The fibrillation kinetics of freshly prepared amylin monomers was accelerated by adding preformed sonicated amylin amyloid seeds as described in the Experimental section. As shown in Fig. 6a (top trace), the addition of amylin monomers to pre-incubated amylin amyloid seeds (at t = 0 s) induced a steady growth of fibers (Movie S1†). The seeding reaction (or fibril growth) and de novo nucleation (binding of additional fibril seeds to the stage) were observed (Fig. 6a). At 1000 s, we observed substantial growth in the fiber morphology; de novo nucleated amylin fibers and several fibers stopped growing after encountering other fibers at their growing end (Fig. 6a and Movie S1†). The growth of the selected amylin fiber (Fig. 6a, rectangle) in the absence of PMAQA is represented as kymographs as a function of time in one direction. The kymograph showed a substantial growth on the free end of the selected amylin fiber (Fig. 6b). These observations agree with the reported observations for other amyloidogenic peptides including β-amyloid.71,72 Remarkably, in the presence of equimolar PMAQA, both amylin fibril growth and de novo nucleation were found to be blocked (Fig. 6c and Movie S2†). In addition, no significant difference in the fiber length and population was observed between 0 s and 1350 s (Fig. 6c). Kymographs of amylin fibrillation in the presence of PMAQA showed no fiber growth. This indicates that PMAQA blocks the recruitment of amylin monomers to the growing ends of amyloid seeds that is needed to proceed the seeding reaction (Fig. 6d). These distinct morphological features observed for both bare mica and APTES-mica suggest that the electrostatic interaction between the sample and the stage did not affect amylin aggregation and interactions with PMAQA. The HS-AFM experimental results, therefore, revealed the inhibitory activity of PMAQA towards amylin aggregation at a molecular-level, which is consistent with other experimental results reported in this study.
Fig. 6 Real-time monitoring of amylin fibrillation using HS-AFM. HS-AFM images show the seeding reaction (fibril growth) of 5 μM amylin monomers in the absence (a) or presence (c) of equimolar PMAQA. These images were extracted at the indicated time intervals from movies S1 and S2 (included in the ESI†). The growth of fibers as a function of time is indicated by arrows. The de novo nucleated amylin fibrils in (a) are shown inside the yellow circles. The kymographs for an individual selected fiber (shown inside pink rectangles in a and c) are shown for amylin in the absence (b) and presence of PMAQA (d). The kymographs show the growth of amylin fibers as a function of time (indicated with pink arrows) in the absence of PMAQA, whereas PMAQA blocks amylin nucleation and seeding processes. The scale bar is 200 nm. |
We further characterized the morphology of β-amyloid-1-40 and amylin species present under various sample conditions (see the Experimental section). The TEM image of 5 μM PMAQA showed very small particles (Fig. 7a), whereas freshly dissolved β-amyloid-1-40 monomers (5 μM) after continuous agitation at room temperature for 24 hours showed a fibril morphology (Fig. 7b). On the other hand, β-amyloid-1-40 monomers (5 μM) incubated with an equimolar concentration of PMAQA exhibited amorphous-like aggregates (Fig. 7c). This correlates with the observed ThT fluorescence results that showed a substantial reduction in fluorescence intensity (Fig. 1a). We next examined the effect of PMAQA on β-amyloid-1-40 aggregation in the presence of amyloid seeds. TEM confirmed the formation of short fibrils when sonicating matured fibers for 2 hours (Fig. 7d). We did not observe a significant change in the sonicated fiber morphology incubated with equimolar PMAQA for 24 hours in the absence of peptide monomers (Fig. 7e). However, a remarkable difference in the β-amyloid-1-40 morphology was observed when a self-seeding reaction was performed in the absence or presence of PMAQA (Fig. 7f and g). The self-seeding reaction presented amyloid fiber growth in the absence of PMAQA (Fig. 7f), whereas in the presence of PMAQA short fibers and an amorphous-like fiber morphology were identified (Fig. 7g). Unlike β-amyloid-1-40, amylin monomers (5 μM) incubated with equimolar PMAQA exhibited no fibers (Fig. 7i) which correlates with the ThT and CD results (Fig. 1b and 2b). Moreover, sonicated amylin fibers incubated with 5 μM PMAQA for 24 hours in the absence of monomers showed a relatively thick fibril morphology as compared to the short-sized sonicated fibers (Fig. 7j–l). Interestingly, sonicated amylin fibers in the presence of equimolar peptide monomers and PMAQA showed no fibril growth which correlates with the HS-AFM observations (Fig. 7k). A quantitative analysis of the widths of sonicated fibers (Fig. 7j) using the ImageJ plugin presented an average fiber width of ∼10.79 nm. In the presence of PMAQA, the sonicated fibers depicted an average width of ∼15.69 nm indicating the possibility of PMAQA binding (Fig. 7l). Overall, the TEM analysis showed the formation of distinct morphological species of β-amyloid-1-40 and amylin in the presence or absence of PMAQA. These results further highlighted the opposite behaviours of PMAQA in the aggregation of β-amyloid-1-40 and amylin as observed in other biophysical experiments.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8sc05771k |
This journal is © The Royal Society of Chemistry 2019 |