Inhibition of Human and Bovine Insulin Fibril Formation by Designed

Article pubs.acs.org/molecularpharmaceutics

Inhibition of Human and Bovine Insulin Fibril Formation by Designed Peptide Conjugates Narendra Kumar Mishra, Khashti Ballabh Joshi, and Sandeep Verma* Department of Chemistry, DST Thematic Unit of Excellence on Soft Nanofabrication, Indian Institute of Technology Kanpur, Kanpur-208016 (UP), India S Supporting Information *

ABSTRACT: The aggregation of insulin, to afford amyloidogenic fibers, is a well-studied phenomenon, which has interesting biological ramifications and pharmaceutical implications. These fibers have been ascribed an intriguing role in certain disease states and stability of pharmaceutical formulations of this hormone. The present study describes the design and inhibitory effects of novel peptide conjugates toward fibrillation of insulin as investigated by thioflavin T assay, circular dichroism (CD), and atomic force microscopy (AFM). Possible interaction of insulin with peptide-based fibrillation inhibitors is also probed by other solution phase studies, which reveal an important role of aromatic π−π interactions in the inhibition process. CD studies suggest that a freshly prepared solution of insulin, rich in α-helices, transforms into a β-sheet structure upon aggregation, which gets perturbed in the presence of synthesized inhibitors. Therefore, these newly designed peptides could serve as potential leads as inhibitors of insulin aggregation. KEYWORDS: insulin, peptide, microscopy, spectroscopy, aggregation, inhibition, NMR β-cells.15−19 An intriguing role of insulin fibers and aggregates, as judged by autoimmune response, has been described in context of Parkinson’s disease.19 It is known that insulin fiber formation occurs through a series of sequential events such as monomer nucleation, elongation, and maturation, which involve an interplay of hydrophobic and electrostatic interactions and the occurrence of cross β-structure.20−24 Insulin aggregates also exhibit different structural isoforms such as spherulites and protofibrils, with interesting optical properties.25−31 It is also known that factors such as temperature, pH, ionic strength, and solvent effects hasten the process of insulin fiber formation.16,21,31 As the understanding of insulin aggregation mechanism and key noncovalent interactions has biochemical ramifications, there are sporadic efforts to study potential inhibitors of this process in the form of organic fluorogens, short peptide sequences, and aromatic compounds.32−39 Notably, the requirement of hydrophobic and electrostatic interactions for insulin aggregation has been identified to aid the design of prospective inhibitors.40 Solvent-assisted pressure tuning experiments revealed the critical role of these interactions in stabilization of the early phase of aggregation. With this available background, we started working on the design of a biocompatible small molecule that would combine hydrophobicity and polar/charged character, for the inhibition of insulin aggregation.

1. INTRODUCTION Several proteins exhibit a tendency to follow molecular mechanisms of amyloidogenesis, which includes conformational changes to acquire β-sheet structural motif, eventually culminating into a number of disease states.1−6 This situation underlies the debilitating role of switch in secondary structural signatures, altered conformational states, and misfolding pathways. Given the inextricable involvement of protein misfolding and aggregation in many neurodegenerative diseases, studies concerning molecular mechanisms of this process have been thoroughly investigated, through a combination of biophysical techniques and functional models. In addition to numerous structural studies concerning amyloid aggregates, considerable efforts are also invested in understanding the molecular basis of cytotoxicity and the degenerative tendency associated with amyloidosis. The ability of amyloid fibrils to elicit a cellular response could be correlated to physicochemical nature of fibrils including features such as length and overall surface area. In this context, some studies have tried to correspond the length of amyloid fibers vis-à-vis its propensity to exhibit cell toxicityit was observed that shorter fibrils have higher cytotoxicity when compared to longer fibers, while some studies also suggest a crucial role of platelet-activating factor as a possible mediator of amyloid cytotoxicity, besides other possible biochemical ramifications.7−14 As these inferences have significant pathophysiological consequences, protein amyloidosis and its ensuing toxicity have received much attention. It is interesting to note that insulin, a peptide hormone crucial for glucose metabolism, exhibits self-association to reveal formation of amyloidogenic aggregates. These aggregates present low levels of cellular toxicity in rat pheochromocytoma and pancreatic © 2013 American Chemical Society

Received: Revised: Accepted: Published: 3903

June 24, 2013 August 26, 2013 September 11, 2013 September 11, 2013 dx.doi.org/10.1021/mp400364w | Mol. Pharmaceutics 2013, 10, 3903−3912

Molecular Pharmaceutics

Article

hydrochloride, DCC, and AMBERLITE IR120 were purchased from Spectrochem, Mumbai, India and used without further purification. 1H and 13C NMR spectra were recorded on JEOLJNM LAMBDA 500 model operating at 500 and 125 MHz, respectively. High-resolution mass spectra (HRMS) were recorded at IIT Kanpur, India, on Waters, Q-Tof Premier micromass HAB 213 mass spectrometer using a capillary voltage of 2.6−3.2 kV. 2.2. Peptide Synthesis. All peptide conjugates were synthesized by solution phase, fragment condensation methodologies using t-Boc chemistry and in the presence of HOBt. The purity of insulin and other products (Figure S1, Supporting Information) was checked by analytical reversephase high-performance liquid chromatography (RP-HPLC) (conditions for all purifications: 0.1% trifluoroacetic acid in water (eluant A) to 0.1% trifluoroacetic acid in acetonitrile (eluant B) with the gradient (0−60% of B in 45 min), flow rate: 1 mL/min, column specification: Agilent’s Eclipse XDB-C18, 4.6 × 250 mm at RT). The purity and identity of all peptide conjugates were confirmed by HRMS. The concentration of all peptide conjugates for a typical analytical run was ∼1 mg/mL (Figure S1). 2.2.1. Synthesis of N-tert-Butyloxycarbonyl-L-tryptophan, N-tert-Butyloxycarbonyl-di-L-trytophan Methyl Ester, and Ntert-Butyloxycarbonyl-di-L-trytophan. These peptide conjugates were synthesized via using standard protocols,43−45 followed by purification and characterization (Scheme S1). 2.2.2. N-tert-Butyloxycarbonyl L-tryptophan-NHS Ester. N-tert-Butyloxycarbonyl L-tryptophan (2.0 g, 6.57 mmol) and N-hydroxysuccinimide (756 mg, 6.57 mmol) were dissolved in a mixture of dichloromethane (4 mL) and dimethylformamide (DMF, 4 mL), and the reaction mixture was cooled to 0 °C under nitrogen atmosphere. Solution of N,N-dicyclohexylcarbodiimide (1.6 g, 7.884 mmol) in dichloromethane (10 mL) was added to the reaction mixture and allowed to stir for 4 h at 0 °C, followed by overnight stirring at the room tempertaure. The reaction mixture was filtered off and filtrate concentrated under reduced pressure. The crude product was diluted with dichloromethane and washed with 1 N HCl (3 × 50 mL), followed by sodium bicarbonate solution (3 × 50 mL) and finally with saturated brine solution (3 × 50 mL). Organic layer was dried over anhydrous sodium sulfate and concentrated to give crude product (yield: 2.2 g, 90%). The crude product was directly used for the next step. 2.2.3. L-Tryptophan-taurine (1). To a stirred solution of NHS ester of N-tert-butyloxycarbonyl L-tryptophan (2.0 g, 4.98 mmol) in 1,4-dioxane (10 mL), taurine (619 mg, 4.98 mmol) and sodium bicarbonate (837 mg, 9.96 mmol), dissolved in water (10 mL), were added, and the reaction mixture was allowed to stir at room temperature for 24 h. The reaction mixture was passed through Amberlite IR 120, and the eluted solution was evaporated to dryness to obtain the crude product. The latter was further purified by silica gel column chromatrography by dichloromethane−methanol (86:14) to get the pure product (1). An Rf value of 0.3 in 20% methanol/dichloromethane was obtained. Yield: 1.2 g (80%). M.P. 130−132 °C; [α]tD = +29.6 (c 0.833 in CH3OH); 1H NMR: (500 MHz, DMSO-d6, 25 °C, TMS) δ(ppm) = 2.50−2.51 (2H, merged with DMSO-d6 signal), 2.92−2.96 (m, 2H), 3.11−3.15 (m, 2H), 3.72−3.75 (t, 1H, J = 6 Hz), 6.94−6.97 (t, 1H, J = 7.45 Hz), 7.03−7.06 (t, 1H, J = 7.45 Hz), 7.1398 (s, 1H), 7.30−7.32 (d, 1H, J = 8.3 Hz) 7.57−7.58 (d, 1H, J = 8.05 Hz), 8.33−8.35 (t, 1H, amide −NH), 10.94 (1H, indolic −NH). 13C NMR (125 MHz,

Insulin consists of 21-residues in A chain and 30 residues in the B chain, which are tethered by one intrachain and two interchain disulfide bonds. During the formation of insulin dimers, a step crucial for amyloidogenesis, the monomer surface assisting dimerization is almost planar and mainly contains nonpolar residues (B22−B30). It has been proposed that three insulin dimers serve as fibril precursors to support the process of insulin aggregation.41 A combined solution phase NMR and MD simulation study with des-Phe(B25) insulin confirmed an altered conformational signature compared to wild-type insulin, leading to loss of aggregation.42 This study suggested the critical role of Phe(B25) and hydrophobic interactions. It was proposed that deletion of Phe(B25) residue moves Pro(B28) residue to B27, thereby altering its interaction with other hydrophobic residues, Val(B12) and Leu(B15), and buries the resultant hydrophobic surface in a way that it no longer supports aggregation. It is also important to note at this point that residues F(B24)−F(B25)−Y(B26) are in the β-strand region and support antiparallel β-sheet formation in the dimer. Finally, looking at the available data on insulin conformation and the possible design combinations, we started off with a Trp-Trp-Tau tripeptide to cover possible hydrophobic and electrostatic considerations in insulin aggregation. Short peptide sequences and their structural aspects are within the ambit of our continuing interest in peptide-based self-assembled aggregates.43−47 In this paper, we decided to explore interaction of two designed peptides (1 and 3) and their dimeric conjugates (2 and 4) comprising of tryptophan and taurine, with human insulin (HI) and bovine insulin (BI) fibrillation. We describe peptide synthesis and their C2symmetrical conjugates (Scheme 1), followed by their ability Scheme 1. Chemical Structures of Water-Soluble Tryptophan and Taurine Containing Inhibitors of Insulin Aggregation

to interfere with insulin aggregation as studied by circular dichroism, fluorescence, atomic force microscopy, and NMR study.

2. MATERIALS AND METHODS 2.1. General. Human insulin was kindly provided by Biocon India Pvt. Ltd. through Dr. Surajit Ghosh, and insulin from Bovine pancreas was purchased from Sigma Aldrich, and thioflavin T from Fluka. Dichloromethane, N,N-dimethylformamide, methanol, triethylamine, and 1,4-dioxane were distilled prior to use following standard procedures. Trifluoroacetic acid, hydrochloric acid, N-hydroxybenzotriazole, t-butyloxycarbonyl carbonate, sodium hydroxide, diethylether, L-amino acids, adipic acid, taurine, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide 3904

dx.doi.org/10.1021/mp400364w | Mol. Pharmaceutics 2013, 10, 3903−3912


Article

DMSO-d6, 25 °C, TMS): δ(ppm) = 28.63, 36.33, 50.96, 54.30, 108.35, 111.96, 118.74, 121.33, 125.16, 127.66, 136.79, 169.95. HRMS-ESI: m/z calcd for C13H18N3O14S [M + H]+ = 312.1018; found: 312.1017. 2.2.4. N-tert-Butyloxycarbonyl-di-L-tryptophan-NHS Ester. N-tert-Butyloxycarbonyl di-L-tryptophan (2.0 g, 4.07 mmol) and N-hydroxysuccinimide (563 mg, 4.89 mmol) were dissolved in a mixture of dichloromethane (4 mL) and DMF (4 mL), and the reaction mixture was cooled to 0 °C under nitrogen atmosphere. A solution of N,N-dicyclohexylcarbodiimide (1.0 g, 4.89 mmol) in dichloromethane (10 mL) was added to the reaction mixture and allowed to stir for 4 h at 0 °C under N2 atmosphere, followed by overnight stirring at the room temperature. The reaction mixture was filtered off, and the filtrate was concentrated under reduced pressure. The crude product was diluted with dichloromethane and washed with 1 N HCl (3 × 50 mL) followed by sodium bicarbonate solution (3 × 50 mL) and further with saturated brine solution (3 × 50 mL). Organic layer was dried over anhydrous sodium sulfate and concentrated to give a crude product (yield: 2.15 g, 90%). The crude was used in the next step without further purification. 2.2.5. Adipic Acid Bis NHS Ester. To a stirred solution of adipic acid (500 mg, 3.421 mmol) in DMF (10 mL) at 0 °C, N-hydroxysuccinimide (1.574 g, 13.68 mmol) and 1-ethyl-3(3dimethylminopropyl)carbodiimide (2.6 g, 13.68 mmol) were added, and the reaction mixture was allowed to stir at room temperature for 24 h. The reaction mixture was evaporated under reduced pressure, and crude product mixture was dissolved in acetone (50 mL) and poured in 1 N HCl (200 mL). This mixture was set aside at room temperature for 1−2 h, and the white precipitate appeared. It was washed with distilled water (2 × 50 mL) followed by hot isopropanol (50 mL). The precipitate was dried under high vacuum to get pure product. Yield: 1.050 g (90%). 1H NMR: (500 MHz, DMSO-d6, 25 °C, TMS) δ(ppm) 1.66−1.69 (m, 4H); 2.69−2.72 (t, 4H, J = 6.6 Hz); 2.77 (s, 8H); 13C NMR (125 MHz, DMSO-d6, 25 °C, TMS): δ (ppm) = 28.56, 30.74, 34.96, 174.09, 175.55. 2.2.6. N,N-Bis-(L-tryptophan-taurine) (2). Bis NHS ester of adipic acid (358 mg, 1.05 mmol) was dissolved in 1,4-dioxane (5 mL), followed by a solution of L-tryptophan-taurine (653 mg, 1.762 mmol) and sodium bicarbonate (176 mg, 2.10 mmol) in water to the reaction mixture. It was allowed to stir at room temperature for 24 h under N2 atmosphere. After completion, the reaction mixture was passed through Amberlite IR 120, and the eluted solution was evaporated to dryness to obtain crude product. It was further purified by silica gel column chromatography using dichloromethane−methanol (80:20), to get pure product 2. Rf = 0.3 in 30% methanol/dichloromethane was obtained. Yield: 515 mg (67%), M.P. 120.0−123.0 °C; [α]tD = −17.46 (c 0.83 in CH3OH); 1H NMR: (500 MHz, DMSO-d6, 25 °C, TMS): δ(ppm) 1.23 (bs, 4H); 1.96 (bs 4H); 2.48 (4H, merged with DMSO-d6 signal); 2.82−2.85 (m, 2H) 3.06−3.13 (m, 4H); 3.30−3.84 (m, 4H) 4.34−4.35 (m, 2H, Trp α,α′-H); 6.90−6.93 (t, 2H, J = 9.75 Hz, Ar-Trp-H); 6.98−7.01 (t, 2H, J = 8.6 Ar-Trp H); 7.06 (s, 2H, Ar-Trp H); 7.25−7.27 (d, 2H, J = 9.75 Hz, Ar-Trp-H); 7.50−7.51 (d, 2H, J = 9.75 Hz, Ar-TrpH); 7.95−7.96 (4H, m, Ar-Trp-H); 10.73(s, 2H indolic −NH). 13 C NMR (125 MHz, DMSO-d6, 25 °C, TMS): δ (ppm) = 25.04, 28.21, 35.36, 35.99, 49.16, 50.81, 54.15, 110.88, 111.81, 118.68, 118.84, 121.30, 124.03, 127.76, 136.56, 171.76, 172.59 . HRMS-ESI: m/z calcd for C32H41N6O10S2 [M + H]+ = 733.2326; found: 733.2320.

2.2.7. Di-L-tryptophan-taurine (3). To a stirred solution of N-tert-butyloxycarbonyl-di-L-tryptophan-NHS ester (2.0 g, 3.40 mmol) in 1,4-dioxane, and to it a solution of taurine (425 mg, 3.40 mmol) and sodium bicarbonate (571 mg, 6.80 mmol) in water was added. The reaction mixure was allowed to stir at room temperature under N2 atmosphere for 24 h. After completion of the reaction, the crude was passed through AMBERLITE IR120. The eluted solution was evaporated to dryness, and the crude compound was further purified by silica gel column chromatography by using dichloromethane− methanol (80:20) to give pure product 3. Rf = 0.3 in 30% methanol/dichloromethane was obtained. Yield: 1.1 g (65%). M.P. 190.0−192.0 °C; [α]tD = −9.639 (c 0.833 in CH3OH); 1 H NMR: (500 MHz, DMSO-d6, 25 °C, TMS) δ(ppm) = 2.51−2.52 (2H, merged with DMSO-d6 signal), 2.95−3.0 (m, 1H) 3.11−3.15 (m, 2H); 3.17−3.21(m, 2H), 3.25−3.28 (2H, merged with DMSO-d6 residual water signal) 3.87−3.90 (m, 1H, Trp αH); 4.38−4.41 (1H, Trp α′-H); 6.92−6.98 (m, 2H, Ar-Trp-H); 7.00−7.05 (m, 2H, Ar-Trp-H); 7.08 (s, 1H, ArTrp-H); 7.13 (s, 1H, Ar-Trp-H); 7.28−7.33 (m, 2H, Ar-TrpH); 7.53−7.55 (d, 1H, J = 7.95 Hz, Ar-Trp-H); 7.64−7.65 (d, 1H, J = 7.95 Hz, Ar-Trp-H); 7.97−7.99 (t, 2H, J = 5.5 Hz, −NH2); 8.80−8.81 (d, 2H, J = 7.05 Hz, amide −NH); 10.80 (s, 1H, indolic −NH); 10.96 (s, 1H, indolic −NH). 13C NMR (125 MHz, DMSO-d6, 25 °C, TMS): δ(ppm) = 27.70, 28.44, 36.24, 48.87, 50.72, 53.33, 54.62, 107.93, 110.35, 111.65, 118.34, 119.08, 121.50, 124.10, 125.39, 127.63, 136.73, 170.18, 170.87. HRMS-ESI: m/z calcd for C24H28N5O5 [M + H]+: 498.1811, found: 498.1812. 2.2.8. N, N-Bis-(Di-L-tryptophan-taurine) (4). Bis NHS ester of adipic acid (300 mg, 0.881 mmol) was dissolved in 1,4dioxane (10 mL), and to it a solution of compound 3 (875 mg, 1.76 mmol) and sodium bicarbonate (142 mg, 1.76 mmol) in water was addded. The reaction mixture was allowed to stir at room temperature under N2 atmosphare for 24 h. After completion of the reaction, the crude product was passed through Amberlite IR 120, and the eluted solution was evaporated to dryness to get the crude product. It was further purified with silica gel column chromatography by dichloromethane−methanol (70:30) to obtain the pure product 4. A Rf value of 0.4 in 40% methanol/dichloromethane was obtained. Yield: 695 mg (69%). M.P. 208−210 °C; [α]tD = −28.91 (c 0.833 in CH3OH); 1H NMR(500 MHz, DMSO-d6, 25 °C, TMS) δ(ppm) = 1.19 (bs, 4H); 1.92 (bs, 4H) ; 2.50 (m, 4H, merged with DMSO-d6 signal); 2.82−2.87 (m, 4H); 2.94−3.02 (m, 4H); 3.05−3.08 (m, 4H); 3.12 (bs, 4H); 4.34−435 (m, 2H); 4.44−4.45 (m, 2H); 6.90−6.91 (m, 2H, Ar-Trp-H), 6.97−7.00 (m, 2H, Ar-Trp-H); 7.05 (bs, 4H, Ar-Trp-H); 7.24− 7.26 (m, 4H, Ar-Trp-H); 7.46−7.47 (d, 2H, J = 7.75 Hz, ArTrp-H); 7.52−7.53 (d, 2H, J = 7.75 Hz, Ar-Trp-H); 7.75−7.77 (m, 2H, amide −NH), 7.88−7.89 (d, 2H, J = 5.0 Hz, amide −NH); 7.95−7.97 (d, 2H, J = 10 Hz amide −NH); 10.73 (s, 2H, indolic −NH); 10.79 (s, 2H, indolic −NH). 13C NMR (125 MHz, DMSO-d6, 25 °C, TMS): δ(ppm) = 25.18, 27.89, 28.06, 35.40, 36.18, 50.94, 54.12, 54.24, 110.42, 110.70, 111.79, 118.65, 118.71, 118.81, 119.00, 121.30, 124.10, 124.21, 127.85, 136.54, 171.24, 172.02, 172.79. HRMS-ESI: m/z calcd for C54H59N10O12S2 [M − H]−: 1103.3761, found: 1103.3826. 2.3. Atomic Force Microscopy (AFM). Fresh and aged incubated samples of insulin alone and mixed with 1, 2, 3, and 4 were imaged with an atomic force microscope (Molecular Imaging, USA) operating under the Acoustic AC mode (AAC), with the aid of a cantilever (NSC 12(c) from MikroMasch). 3905



Article

The force constant was 0.6 N/m, while the resonant frequency was 150 kHz. The images were taken in air at room temperature, with the scan speed of 1.5−2.2 lines/s. The data acquisition was done using PicoScan 5 software, while the data analysis was done using of visual SPM. A portion of 10 μL of fresh and incubated samples was diluted by 100 μL of 0.1 N HCl water (pH 1.6), and out of it 5 μL was deposited onto a freshly cleaved mica surface at room temperature. The sample was uniformly spread using a spin-coater operating at 200−500 rpm (PRS-4000). The sample-coated mica was dried at room temperature in a dust-free space for 30 min followed by AFM imaging. 2.4. Thioflavin T (ThT) Binding Assay. To a freshly prepared solution of insulin alone (170 μM) or incubated with compound 1, 2, 3, and 4 (100 μM) and ThT (20 μL, 1 mM) were added to a final ThT concentration of 20 μM. These samples were allowed to incubate at 65 °C, and the fluorescence intensity was measured at room temperature (λex = 410 nm and λem = 488 nm). Fluorescence spectra were recorded on a Varian Luminescence Cary Eclipse with a 10 mm quartz cell at 65 ± 0.1 °C using a Peltier element. Slits (bandpass) were set to 5 nm, and samples were magnetically stirred at 800 rpm to achieve homogeneous excitation. HPLC grade water and AR grade hydrochloric acid were used. 2.5. Circular Dichroism Spectroscopy. A freshly prepared stock solution of insulin (170 μM) was either incubated alone or with conjugates 1, 2, 3, and 4 (100 μM,) at pH 1.6 (65 °C). All experiments were carried out at room temperature, and spectra were collected for a final concentration of insulin (17 μM), at different time intervals using JASCO J-815 CD spectrometer and a quartz cuvette, with a path length of 1 mm. CD spectra were collected between 195 to 270 nm, and each spectrum was the average of three scans. To avoid any instrumental baseline drift between any measurements, the background value was subtracted for each individual sample measurement with (0.1 N HCl, pH 1.6). 2.6. NMR Spectroscopy. Samples were prepared by dissolving the lyophilized solution of insulin and pure sample of inhibitor (3 and 4) in 0.1 N HCl, followed by the addition of increasing amounts of DMSO-d6. 30% DMSO-d6/0.1 N HCl solution (pH 1.6) was found to be a good solvent system for the NMR measurements. 1H NMR experiments were carried out at 25 °C on 500 MHz JEOL ECX spectrometer. Each spectrum was recorded for 1000 scans. 2.7. 3D Peptide Structure. The crystallography data used for 3D peptide structure display and modeling were extracted from the PDB file 2ZP6. The interactive visualization and analysis of 3D molecular structures of the insulin hexamer, dimer, and monomer were displayed using PyMolwin and CHIMERA 1.8rc.

Figure 1. AFM micrographs of freshly incubated human and bovine insulin (A, B), respectively; human and bovine insulin, respectively, after 20 h incubation (pH 1.6, 65 °C) (C, D) (scale bar = 1 μm).

to an exponential increase in its fluorescence intensity.49−52 Typically, ThT dye (20 μM) was incubated with insulin under standard aggregation conditions, and an increase in emission was observed with a maxima at ∼488 nm, up to a 20 h duration, suggesting its facile binding to amyloidogenic insulin fibers. Peptides and their conjugates 1−4 were coincubated with insulin, and their effect on aggregation was assessed by ThT staining. Notably, conjugate 4 exhibited the strongest inhibition among all conjugates followed by 3, which was markedly better compared to dipeptide 1 and its dimer conjugate 2 (Figure 2).

3. RESULTS AND DISCUSSION We started our investigations by first establishing insulin incubation and aggregation protocol as per literature described procedures (170 μM, pH 1.6 at 65 °C; 20 h), to confirm the appearance of fibrillar structures (Figure 1).21,48 It is known that a mixture of monomers and dimers of natively folded insulin seed the evolution of amyloidogenic fibers. AFM images (on mica surface) of both HI and BI confirmed reproducible fiber formation after 20 h of incubation. This protocol was employed for all subsequent experiments. Thioflavin T (ThT) is employed for the detection of amyloidogenic protein aggregates as its binding to fibrils leads

Figure 2. Fluorescence screen of BI and HI aggregation in the absence and presence of conjugates 1−4, at λex = 410 nm, λem = 488 nm.

This remarkable effect of 4 and other conjugates on insulin aggregation was further probed with CD spectroscopy and AFM. Human (HI) and bovine insulin (BI) show an α-helix rich CD pattern along with two negative bands centered around 208 and 222 nm, consistent with the gross structure of insulin hexamer (Figure 3).21,48 Time-dependent CD experiments 3906



Article

Figure 3. Far UV-CD spectra for insulin aggregation and its interference by 4 (100 μM). Left: with BI and Right: with HI.

Figure 4. Time-dependent AFM micrographs of coincubation of insulin with 4: with HI at 0 h (A), 20 h (B), and 40 h (C), respectively, and with BI at 0 h (D), 20 h (E), and 40 h (F), respectively (scale bar = 1 μm).

4 at t = 0 min, with both HI and BI in situ, showed the presence of gross spherical morphology, instead of fibers (Figure 4A−F). These spherical structures resemble morphologies formed during the early nucleation phase of the insulin aggregation process (Figure 1A,B).20,26 The presence of 4 at t = 0 min arrests the growth of fibrous structures, over the duration of 40 h for BI and 80 h for HI, and time-dependent AFM studies further confirm that 4 effectively inhibits insulin aggregation perhaps by interfering with the early events of this process (Figures S14−S19). To further evaluate the inhibitory effect of conjugate 4 on insulin fiber formation, an experiment involving time dependence was conducted (Figure 5).35 4 was added to human insulin solution at different time points from t = 0−6 h, with a one hour time interval, and the solutions were further

suggested that the spectral profile changed during the initial incubation phase (fresh to 1 h), while exhibiting a tendency to revert to the original profile for a longer incubation period (see Supporting Information). Interestingly, BI affords a β-sheet rich structure as judged by a peak at ∼218 nm after 60 min of incubation, while HI affords a similar transformation after a much longer incubation of 10 h (Figure 3 and Figure S8). Notably, 4 prevented loss of the α-helix structure for both human and bovine insulin when coincubated from t = 0−80 h (Figure 3 and Figure S12). Peptides and their conjugates 1−3 also followed the trend observed in the ThT assay suggesting that their binding to insulin stabilizes helix-rich native structure, despite favorable conditions available for aggregation (see Supporting Information). We resorted to atomic force microscopy to determine morphological consequences of such interactions. Co-incubation of 3907



Article

of 4 with human insulin.54,55 It is worth mentioning that insulin lacks emission at ∼360 nm, a photophysical property generally associated with Trp emission. Titration of insulin to a solution of 4 lead to a gradual reduction in fluorescence intensity perhaps due to hydrophobic interactions between 4 and amino acid residues present in insulin B chain (Figure S20). In a control experiment, tyrosine alone was titrated with the inhibitor 4, and a similar fluorescence quenching was observed (Figure S21), suggesting possibility of Tyr-Trp interaction. However, inhibitory contribution of other possible amino acid side chains cannot be ruled out at present. The extended C-terminal ends of the two insulin chains (A and B chains) support the formation of two-stranded antiparallel β-sheets, which is further stabilized by hydrophobic contacts and hydrogen bonds during dimer formation.56,57 It is also important to note that the deletion of Phe(B25) destabilizes hydrophobic contact causing interference with dimer formation.57,42 On the basis of fluorescence studies, it could be surmised that a hydrophobic tripeptide motif (Phe-Phe-Tyr) present in the B-chain could possibly interact with the designed inhibitor, thereby preventing dimer-stabilizing hydrophobic contacts (Figure 7). To investigate this issue, we carried out solution state 1H NMR study of insulin, inhibitors 3 and 4 alone, and inhibitors with insulin in 30% DMSO-d6 in 0.1 N HCl (pH 1.6; 25 °C), to maintain amyloidogenic conditions. 1H NMR analysis offered interesting insight into the interaction. Inhibitor− insulin interaction caused a significant downfield shift (marked with dotted line and star) in inhibitor aromatic, amide (−NH), and indolic (−NH) protons followed by slightly upfield shift with peak broadening of aromatic (specially Phe-Phe-Tyr) protons of insulin, which is interpreted as an effect exerted by ring current shifts induced by π−π stacking between inhibitor and insulin (Figure 8).58,59 Different aggregative forms of insulin are known to exist in solution, which depend on physicochemical factors such as solvent, concentration, pH, metal ions, and on its interaction with small molecules.60−62 In this context the small aromatic molecules are well-known to inhibit fibril formation of amyloidogenic proteins owing to their aromatic stacking ability. Tyr(B16) also plays a pivotal role during the assembly process.35 Moreover, it is known that many aromatic amino

Figure 5. Inhibitory effect of conjugate 4 on time-dependent fibril formation of insulin. 4 (100 μM) was added to human insulin solution (170 μM, pH 1.6, 65 °C) incubated at various time intervals. Fluorescence intensity measured for ThT: λex = 410 nm, λem = 488 nm.

incubated up to 30 h. The highest inhibition was observed with 4 at 0 h, when incubated at 65 °C. Addition of 4 after 4 h and beyond did not affect fluorescence intensity to an appreciable extent. These results indicate that 4 binds to insulin in the early nucleation phase, which is in agreement with our AFM results. In this context, it is interesting to note that Bucciantini and coworkers have showed that amyloid species formed during the early phases of aggregation can be highly cytotoxic in nature.7 Thus, it could be surmised that 4 will be effective in interfering with species that appear during the early phase of aggregation. To resolve the effect of inhibitor concentration on the lack of fibrillation, conjugate 4 was subjected to a concentrationdependent fluorescence assay. ThT fluorescence assay confirmed a visible correlation of concentration with extent of inhibition, over a 20 h time period, for both bovine and human insulin (Figure 6). Notably, the inhibitory effect of 4 at 12.5 μM for HI was considerably more pronounced compared to BI, suggesting its better efficacy toward the human hormone. Tryptophan fluorescence quenching, in human apolipoprotein A-IV, occurs due to Tyr-Trp interaction, thereby providing crucial information concerning location of the amino terminus.53 As insulin consists Tyr residues, we decided to follow changes in Trp fluorescence intensity due to interaction

Figure 6. Concentration-dependent inhibitory effect of 4 after 20 h incubation with human and bovine insulin. 3908



Article

Figure 7. Dimeric structure of bovine insulin (PDB ID: 2ZP6) (A) Top: dimeric structure with a side chain showing the presence of Phe-Phe-Tyr residue is the predicted interaction site (bottom: FFY in B-chain).

Figure 8. 500 MHz 1H NMR spectra of amide/aromatic region of (A) inhibitor 3, human insulin, and inhibitor 3 + insulin and (B) inhibitor 4, insulin, and 4 + insulin in 30% DMSO-d6−0.1 N HCl (pH 1.6). NMR showing the significant downfield shift of indolic, amidic, and aromatic protons for the inhibitor and the interaction of aromatic protons of insulin with the aromatic protons of the inhibitor (marked with dotted box), followed by peak broadening in insulin aromatic protons (Figure S23).

acid containing peptides63 inhibit insulin fibril formation, suggesting that small aromatic molecules could be very

promising candidates for modifying amyloidogenic insulin fibrillation. 3909



Article

Figure 9. Proposed model for insulin (A) aggregation and (B) inhibition by peptide conjugates. Path A showing that the mechanism of insulin fiber formation,20,35 while path B shows the proposed mechanism of inhibition of insulin fibrillation through binding of conjugate 4 to the aromatic-rich region of insulin (pathway adapted from ref 35).

Notes

Therefore, based on the above results the predicted binding location of inhibitor is in the aromatic-rich region of insulin B-chain (for example, tripeptide motif from B24−26). Figure 9 describes a tentative model of binding and inhibition of insulin fiber formation in the early phase of aggregation by designed conjugate 4.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS N.K.M. thanks CSIR-India, for predoctoral research fellowship. K.B.J. thanks DST for financial support. This work is supported by Outstanding Investigator Award to S.V. from DAE-SRC, Department of Atomic Energy, India and by DST through J. C. Bose National Fellowship.

4. CONCLUSION In conclusion, we have designed short peptide sequences and their conjugates, based on some principles implicated for insulin aggregation, and have investigated their ability to interfere and inhibit formation of insulin amyloidogenic fibers. This behavior was studied with the help of ThT assay, circular dichroism, and atomic force microscopy. 1H NMR study was also carried out to investigate possible interaction of insulin with the inhibitor. Concentration-dependent inhibition by conjugate 4 provides an impetus to optimize the design of such molecules that may eventually have relevance in pharmaceutical applications. Future work will also focus on delineating structural requirements in short peptides, making these molecules more stable under the physiological conditions and to see their effect in living systems.

■

■

(1) Selkoe, D. J. Proteins folding in fatal ways. Nature 2003, 426, 900−904. (2) Kelly, J. W. Towards understanding of molecular amyolgenesis. Nat. Struct. Biol. 2002, 9, 323−325. (3) Levine, H. The challenge of inhibiting Aβ polymerization. Curr. Med. Chem. 2002, 9, 1121−1133. (4) Tycko, R.; Wickner, R. B. Molecular structure of amyloid and prions fibrils: Consensus versus controversy. Acc. Chem. Res. 2013, 46, 1487−1496. (5) Sipe, J. D.; Cohen, A. S. Review: History of the amyloid fibril. J. Struct. Biol. 2000, 130, 88−98. (6) Morrisa, K.; Serpell, L. From natural to designer self assembles biopolymer the structural characterization of fibrous proteins and peptides using fibre diffraction. Chem. Soc. Rev. 2010, 39, 3445−3453. (7) Bucciantini, M.; Giannoni, E.; Chiti, F.; Baroni, F.; Formigli, L.; Zurdo, J.; Taddei, N.; Ramponi, G.; Dobson, C. M.; Stefani, M. Inherent toxicity of aggregates implies a common mechanism for protein misfolding diseases. Nature 2002, 416, 507−511. (8) Sirangelo, I.; Irace, G. Inhibition of aggregate formation as therapeutic target in protein misfolding diseases: effect of tetracycline and trehalose. Expert Opin. Ther. Targets 2010, 14, 1311−1321. (9) Sirangelo, I.; Iannuzzi, C.; Vilasi, S.; Irace, G.; Giuberti, G.; Misso, G.; D’Alessandro, A.; Abbrezzese, A.; Caraglia, M. W7FW14F Apomyoglobin amyloid aggregates-mediated apoptosis is due to oxidative stress and AKT inactivation caused by Ras and Rac. J. Cell Physiol. 2009, 221, 412−423. (10) Vilasi, A.; Vilasi, S.; Romano, R.; Acernese, F.; Barone, F.; Balestrieri, M. L.; Maritato, R.; Irace, G.; Sirangelo, I. Unraveling amyloid toxicity pathway in NIH3T3 cells by a combined proteomic and 1H-NMR metabolomic approach. J. Cell. Physiol. 2013, 228, 1359−1367. (11) Malmo, C.; Vilasi, S.; Iannuzzi, C.; Tacchi, S.; Cametti, C.; Irace, G.; Sirangelo, I. Tetracycline inhibits W7FW14F apomyoglobin fibril

ASSOCIATED CONTENT

S Supporting Information *

Synthesis details, RP-HPLC traces of compounds 1, 2, 3, 4, and HI, HRMS-ESI of HPLC purified compounds 1, 2, 3, and 4, time-dependent fluorescence of BI and HI with ThT, far UVCD spectra for peptides 1−4 and BI and HI, fluorescence of BI and HI, AFM micrographs of conjugates 1−4 and BI and HI fiber formation and inhibition, titration of inhibitor, and NMR spectra. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel.: 91-11-512-259-7643. Fax: 91-11-512-259-7436. Present Address

K.B.J.: Department of Chemistry, Panjab University, Chandigarh-160014 (Panjab), India. 3910



Article

extension and keeps the amyloid protein in a pre-fibrillar, highly cytotoxic state. FASEB J. 2006, 20, 346−347. (12) Grudzielanek, S.; Velkova, A.; Shukla, A.; Smirnovas, V.; Tatarek-Nossol, M.; Rehage, H.; Kapurniotu, A.; Winter, R. Cytotoxicity of insulin within its self-assembly and amyloidogenic pathways. J. Mol. Biol. 2007, 370, 372−384. (13) Wang, S. S.-S.; Liu, K.-N.; Han, T.-C. Amyloid fibrillation and cytotoxicity of insulin are inhibited by the amphiphilic surfactants. Biochim. Biophys. Acta 2010, 1802, 519−530. (14) Kachooei, E.; Moosavi-Movahedi, A. A.; Khodagholi, F.; Ramshini, H.; Shaerzadeh, F.; Sheibani, N. Oligomeric forms of insulin amyloid aggregation disrupt outgrowth and complexity of neuron-like PC12 cells. PloS One 2012, 7, e41344. (15) Waugh, D. F. Regeneration of insulin from insulin fibrils by the action of alkli. J. Am. Chem. Soc. 1948, 70, 1850−1857. (16) Nielsen, L.; Khurana, R.; Coats, R. A.; Frokjaer, S.; Brange, J.; Vyas, S.; Uversky, V. N. Effect of environmental factor on the kinetics of insulin fibril formation: Elucidation of molecular mechanism. Biochemistry 2001, 40, 6036−6046. (17) Ivanova, M. I.; Sievers, S. A.; Sawaya, M. R.; Wall, J. S.; Eisenberg, D. Molecular basis for insulin fibril assembly. Proc. Natl. Acad. Sci. 2009, 106, 18990−18995. (18) Brange, J.; Andersen, L.; Laursen, E. D.; Meyn, G.; Rasmussen, E. Towards understanding of insulin fibrillation. J. Pharm. Sci. 1997, 86, 517−525. (19) Wilhelm, K. R.; Yanamandra, K. M.; Gruden, A.; Zamotin, V.; Malisauskas, C. M.; Darinskas, V. A.; Forsgren, L.; Morozova-Roche, L. A. Immune reactivity towards insulin, its amyloid and protein S100B in blood in sera of Parkinson’s disease patients. Eur. J. Neurol. 2007, 14, 327−334. (20) Jansen, R.; Dzwolak, W.; Winter, R. Amyloidogenic self assembly of insulin aggregates probed by high resolution atomic force microscopy. Biophys. J. 2005, 88, 1344−1353. (21) Nielsen, L.; Frokjaer, S.; Brange, J.; Uversky, V. N.; Fink, A. L. Probing the mechanism of insulin fibril formation with insulin mutants. Biochemistry 2001, 40, 8397−8409. (22) Babenko, V.; Piejko, M.; Wójcik, S.; Mak, P.; Dzwolak, W. Vortex-induced amyloid superstructures of insulin and its component A and B chains. Langmuir 2013, 29, 5271−5278. (23) Librizzi, F.; Rischel, C. The kinetics behavior of insulin fibrillation is determined by heterogeneous nucleation pathways. Protein Sci. 2005, 14, 3129−3134. (24) Burke, M. J.; Rougvie, M. A. Cross β Protein structure. Biochemistry 1972, 11, 2435−2439. (25) Chwedoruk, W. S.; Nieznanska, H.; Wojcik, S.; Dzwolak, W. Cross seeding of fibrils from two types of insulin induced new amyloid strains. Biochemistry 2012, 51, 9460−9469. (26) Krebs, M. R. H.; Maphee, C. E.; Miller, A. F.; Dunlop, I. E.; Dobson, C. M.; Donald, A. M. The formation of spherulites by amyloid fibres of bovine insulin. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 14420−14424. (27) Kurouski, D.; Dukor, R. K.; Lu, X.; Nafie, L. A.; Lednev, I. K. Normal and reversed supramolecular chirality of insulin fibrils probed by vibrational circular dichroism at the protofilament level of fibrils structure. Biophys. J. 2012, 103, 522−531. (28) Fodera, V. M.; Weert, V. D.; Vestergaard, B. Large scale polymorphism and auto-catalytic effect in insulin fibrillogenesis. Soft Matter 2010, 6, 4413−4419. (29) Loksztejn, A.; Dzwolak, W. Chiral bifurcation in aggregating insulin: an induced circular dichroism study. J. Mol. Biol. 2008, 379, 9− 16. (30) Amdursky, N.; Gazit, E.; Rosenman, G. Formation of lowdimensional crystalline nucleus region during insulin amyloidogenesis process. Biochem. Biophys. Res. Commun. 2012, 419, 232−237. (31) Sluzky, V.; Tamada, J. A.; Klibanov, A. M.; Langer, R. Kinetics of insulin aggregation in aqueous solutions upon agitation in the presence of hydrophobic surfaces. Proc. Natl. Acad. Sci. 1991, 88, 9377−9381. (32) Arora, A.; Ha, C.; Park, C. B. Inhibition of insulin amyloid formation by small stress molecules. FEBS Lett. 2004, 564, 121−125.

(33) Vilasi, S.; Lannuzzi, C.; Portaccio, M.; Irace, G.; Sirangelo, I. Effect of trehalose on W7FW14F apomyoglobin and insulin fibrillization: New insights into the inhibition activity. Biochemistry 2008, 47, 1789−1796. (34) Giger, K.; Vanam, R. P.; Seyrek, E.; Dubin, P. L. Suppression of insulin aggregation by Heparin. Biomacromolecules 2008, 9, 2338− 2344. (35) Hong, Y.; Meng, L.; Chen, S.; Leung, C. W. T.; Da, L. T.; Faisal, M.; Silva, D. A.; Liu, J.; Lam, J. W.; Huang, Y. X.; Tang, B. Z. Monitoring and inhibition of insulin fibrillation by small organic fluorogen with aggregation-induced emission characteristics. J. Am. Chem. Soc. 2012, 134, 1680−1689. (36) Gibson, T. J.; Murphy, R. M. Inhibition of insulin fibrillogenesis with targeted peptide. Protein Sci. 2006, 15, 1133−1141. (37) Velkova, A.; Nossol, M. T.; Andreetto, E.; Kapurniotu, A. Exploiting cross-amyloid interactions to inhibit protein aggregation but not function: nanomolar affinity inhibition of insulin aggregation by an IAPP mimic. Angew. Chem., Int. Ed. 2008, 47, 7114−7118. (38) Wang, S. H.; Dong, X. Y.; Sun, Y. Effect of (−) epigallocatechin3-gallate on human insulin fibrillation/aggregation kinetics. Biochem. Eng. J. 2012, 63, 38−49. (39) Varughese, M. M.; Newman, J. Inhibitory effect of arginine on the aggregation of bovine insulin. J. Biophys. 2012, 434289. DOI: 10.1155/2012/434289. (40) Grudzielanek, S.; Smirnovas, V.; Winter, R. Solvent-assisted pressure tuning of insulin fibrillation: from novel aggregation pathways to biotechnological applications. J. Mol. Biol. 2006, 356, 497−509. (41) Nayak, A.; Sorci, M.; Krueger, S.; Belfort, G. A. Universal pathways for amyloid nucleus and precursor formation for insulin. Proteins 2009, 74, 556−565. (42) Jørgensen, A. M. M.; Olsen, H. B.; Bahlschmidt, P.; Led, J. J. Solution structure of the superactive monomeric Des-[Phe(B25)] Human insulin mutants: Elucidation of the structural basis for the monomerization Des-[Phe(B25)] of insulin and the dimerisation of native insulin. J. Mol. Biol. 1996, 257, 684−699. (43) Mondal, S.; Barman, A. K.; Verma, S. Peptide-based synthetic design construction and morphology of soft structures. Chimia 2012, 66, 930−935. (44) Joshi, K. B.; Verma, S. Ditryptophan conjugation triggers conversions of biotin fibres into soft spherical structures. Angew. Chem., Int. Ed. 2008, 47, 2860−2863. (45) Ghosh, S.; Reches, M.; Gazit, E.; Verma, S. Bioinspired design of nanocages by self assembling of triskelion peptide elements. Angew. Chem., Int. Ed. 2007, 46, 2002−2004. (46) Ghosh, S.; Verma, S. Phase fiiber growth in peptide conjugates: aggregation and disaggregation studies. J. Phys. Chem. B. 2007, 111, 3750−3757. (47) Ghosh, S.; Singh, S. K.; Verma, S. Self assembly and potassium ion triggered disruption of peptide based-soft structures. Chem. Commun. 2007, 2296−2298. (48) Bouchard, M.; Zurado, J.; Nettleleton, E. J.; Dobson, C. M.; Robinson, C. V. Formation of insulin amyloid fibrils followed by FTIR simultaneously with CD and electron microscopy. Protein Sci. 2000, 9, 1960−1967. (49) Krebs, M. R. H.; Bromley, E. H. C.; Donald, A. M. The binding of thioflavin-T to amyloid fibrils: localization and implications. J. Struct. Biol. 2005, 149, 30−37. (50) Khurana, R.; Coleman, C. C.; Zanetti, I. S.; Carter, A.; Krishna, V.; Grover, R. K.; Roy, R.; Singh, S. Mechanism of thioflavin-T binding to amyloid fibrils. J. Struct. Biol. 2005, 151, 229−238. (51) Biancalana, M.; Koide, S. Molecular mechanism of thioflavin-t bindings to amyloid fibres. Biochim. Biophys. Acta 2010, 1804, 1405− 1412. (52) Sulatskaya, A. I.; Kuznetsova, I. M.; Turoverov, K. K. Interaction of thioflavin-t with amyloid fibrils: stoichiometry and affinity of dye binding, absorption spectra of bound dye. J. Phys. Chem. B 2011, 115, 11519−11524. 3911



Article

(53) Weinberg, R. B. Exposure and electronic interaction of tyrosine and tryptophan residues in human apolipoprotein A-IV. Biochemistry 1988, 27, 1515−1521. (54) Bekard, I. B.; Dunstan, D. E. Tyrosine auto fluorescence as a measure of bovine insulin fibrillation. Biophys. J. 2009, 97, 2521−2531. (55) Bodenreider, C.; Beer, D.; Keller, T. H.; Sonntag, S.; Wena, D.; Li Jian, Y.; Yau, Y. H.; Shochat, S. G.; Huang, D.; Zhou, T.; Caflisch, A.; Su, X.-C.; Ozawa, K.; Otting, G.; Vasudevan, S. G.; Lescar, J.; Lim, S. P. A fluorescence quenching assay to discriminate between specific and nonspecific inhibitors of dengue virus protease. Anal. Biochem. 2009, 395, 195−204. (56) Su, X.-C.; Ozawa, K.; Yagi, H.; Lim, S. P.; Wen, D.; Ekonomiuk, D.; Huang, D.; Keller, T. H.; Sonntag, S.; Caflisch, A.; Vasudevan, S. G.; Otting, G. NMR study of the complexes between low molecular mass inhibitors and the west line Virus NS2B-NS3 protease. FEBS J. 2009, 276, 4244−4255. (57) Keller, D.; Clausen, R.; Josefsen, K.; Led, J. J. Flexibility and bioactivity of insulin: an NMR investigation of the solution structure and folding of an unusually flexible human insulin mutant with increased biological activity. Biochemistry 2001, 40, 10732−10740. (58) Mahalakshmi, R.; Raghothama, S.; Balaram, P. NMR analysis of aromatic interaction in designed peptide β-hairpins. J. Am. Chem. Soc. 2006, 128, 1125−1138. (59) Coleman, J. E.; Anderson, R. A.; Ratcliffe, R. G.; Armitage, I. M. Structure of gene 5 protein-oloigodeoxynucleotide complexes as a determined by proton 19F, and 31P nuclear magnetic resonance. Biochemistry 1976, 15 (25), 5419−5430. (60) Becker, R. H. Insulin glulisine complementing basal insulins: a review of structure and activity. Diabetes Technol. Ther. 2007, 9, 109− 121. (61) Brange, J.; Langjær, L.; Pearlman, R.; Wang, Y. J., Eds. Insulin structure and stability (Chapter 11). Stability and Characterization of Protein and Peptide Drugs: Case Histories; Springer: New York, 1993; p 315. (62) Levy-Sakin, M.; Shreberk, M.; Daniel, Y.; Gazit, E. Targeting insulin amyloid assembly by small aromatic molecule: towards rational design of aggregation inhibitors. Islets 2009, 1, 210−215. (63) Wu, C.; Lei, H.; Duan, Y. The role of Phe in the formation well ordered oligomers of amyloidogenic hexapeptide (NFGAIL) observed in molecular dynamics simulations with explicit solvent. Biophys. J. 2005, 88, 2897−2906.

3912
