Insight into the Inhibition Effect of Acidulated Serum Albumin on

Aug 1, 2014 - Bifunctionality of Iminodiacetic Acid-Modified Lysozyme on Inhibiting Zn-Mediated Amyloid β-Protein Aggregation. Xi LiBaolong XieXiaoya...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Insight into the Inhibition Effect of Acidulated Serum Albumin on Amyloid β‑Protein Fibrillogenesis and Cytotoxicity Baolong Xie,† Xi Li,† Xiao-Yan Dong,†,‡ and Yan Sun*,†,‡ †

Department of Biochemical Engineering and Key Laboratory of Systems Bioengineering of the Ministry of Education, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China ‡ Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, China S Supporting Information *

ABSTRACT: Alzheimer’s disease (AD) is the most prevalent form of dementia, and aggregation of amyloid β-proteins (Aβ) into soluble oligomers and fibrils has been implicated in the pathogenesis of AD. Herein we developed acidulated serum albumin for the inhibition of Aβ42 fibrillogenesis. Bovine serum albumin (BSA) was modified with diglycolic anhydride, leading to the coupling of 14.5 more negative charges (carboxyl groups) on average on each protein surface. The acidulated BSA (A-BSA) was characterized and confirmed to keep the tertiary structure and stability of BSA. Extensive biophysical and biological analyses showed that A-BSA significantly inhibited Aβ42 fibrillogenesis and mitigated amyloid cytotoxicity. As compared to the Aβ42-treated group (cell viability, 50%), the cell viability increased to 88% by the addition of equimolar A-BSA. The inhibitory effect was remarkably higher than that of BSA at the same concentration. On the basis of the experimental findings, a mechanistic model was proposed. The model considers that Aβ42 is bound to the A-BSA surface by hydrophobic interactions, but the widely distributed negative charges on the A-BSA surface give rise to electrostatic repulsions to the bound Aβ42 that is also negatively charged. The two well-balanced opposite forces make Aβ42 adopt extended conformations instead of the β-sheet structure that is necessary for the on-pathway fibrillogenesis, even when the protein is released off the surface. Thus, A-BSA greatly slows down the fibrillation and changes the fibrillogenesis pathway, leading to the formation of less toxic aggregates. The findings and the mechanistic model offer new insights into the development of more potent inhibitors of Aβ fibrillogenesis and cytotoxicity.

1. INTRODUCTION Alzheimer’s disease (AD) is now recognized as the most prevalent neurodegenerative disease.1,2 This disease leads to the progressive dysfunction of memory, disordered cognitive function, and death ultimately of the patients.3−6 The morbidity of AD increases with age, and more than 35 million people were suffering from AD in 2011. Because of the aging population, AD has become the fifth leading cause of death among the elderly in the USA and has caused huge economic loss.7 Moreover, AD is still a serious illness with no cure.8 Hence, effective prevention and treatment of AD have drawn worldwide attention. AD is characterized by extracellular amyloid plaques and intracellular neurofibrillary tangles in pathology.1,9 The amyloid plaques are mainly composed of the aggregates (i.e., oligomers and fibrils) of amyloid β-proteins (Aβ).10,11 Aβ containing 39− 43 amino acid residues are derived through sequential proteolytic cleavage of amyloid β precursor protein,1,12,13 and the most prevalent variations are Aβ40 and Aβ42, with Aβ42 being the more amyloidogenic form.14 Elevated levels of Aβ lead to the aggregation of Aβ, regardless of soluble oligomers or mature fibrils, which are neurotoxic to brain cells associated © 2014 American Chemical Society

with perturbations of synaptic function and neural network activity that probably underlie the cognitive deficits in AD.15,16 Therefore, the inhibition of Aβ aggregation is one of the keys to the prevention and treatment of AD. At present, a lot of inhibiting agents of Aβ aggregation have been reported, such as peptides,17,18 small organic molecules,19,20 nanoparticles (NPs),21,22 and proteins.23,24 For the effective use of the inhibitors, high biocompatibility in vivo is desired.25 Consequently, it is desired to develop peptide-/ protein-based biomaterials that function on the inhibition of Aβ aggregation. It has been known that some proteins play a significant role in inhibiting the aggregation of Aβ in vivo, and it has been found that the decrease in concentration of these proteins with aging is a risk factor leading to AD.26 Among them, human serum albumin (HSA) has been found to bind Aβ40 and Aβ42 at a 1:1 stoichiometry with a dissociation constant of 5−10 μM.14,27 Because of the high concentration of HSA in blood plasma (640 μM), HSA can bind 90%−95% of Received: June 26, 2014 Revised: August 1, 2014 Published: August 1, 2014 9789

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796

Langmuir

Article

deionized water at 10 mg/mL. The protein solution was kept at 25 °C in a water bath, and diglycolic anhydride powder was slowly added to the protein solution. The reaction mixture was maintained at pH 6.0 by titration with 1.0 M NaOH. After reaction at 25 °C for 2 h, the reactant was loaded onto a Sephadex G25 column for separation by size-exclusion chromatography (SEC) to remove excess diglycolic anhydride and other byproducts. The separation was performed on an Ä KTA FPLC system (GE Healthcare) equipped with an XK16/20 column (16 mm I.D., 20 cm in length) packed with 40 mL of Sephadex G25 gel. The column was washed with 20 mM ammonium acetate buffer (pH 7.4 ± 0.05) at 1 mL/min. The modified protein fraction was collected and freeze-dried under vacuum for 24 h. The lyophilized protein (A-BSA) was stored at −20 °C. A-BSA was analyzed by matrix-assisted laser desorption ionization time-of-flight mass spectroscopy (Autoflex Tof/TofIII, Bruker Daltonics Inc, Billerica, MA, USA) to determine its modification degree of carboxyl groups at 1 mg/mL. The zeta potentials of BSA and A-BSA (1 mg/mL) at different pH values were determined by Zetasizer Nano (Malvern Instruments, Worcestershire, UK) in deionized water, whose pH value was adjusted by adding NaOH or HCl at 25 °C. The molecular sizes and the intrinsic fluorescence spectra of BSA and A-BSA were respectively measured on a Zetasizer Nano and fluorescence spectrometer (Perking Elmer LS-55, MA, USA) in buffer A (10 mM sodium phosphate and 100 mM NaCl, pH 7.4 ± 0.05) in 1 mg/mL solution at 25 °C. The net charge of BSA at pH 7.4 was calculated from the following equation with the primary sequence data of BSA:38

Aβ species in blood plasma.14,28 Thus, HSA acts as an “external sink” of Aβ14,24,29 and plays a significant role in mitigating the cytotoxicity of Aβ oligomers and fibrils in vivo. However, at physiological conditions, HSA has very little effect on the elongation rate of fibrils.14 This means that its inhibition effect on Aβ aggregation is quite limited. The presence of albumin in the human brain has been reported,14,30,31 indicating the penetration capability of HSA across the blood−brain barrier. Hence, if we can enhance the inhibition effect of albumin by chemically modifying its molecular surface, it would function as both the “external sink” of Aβ and the aggregation inhibitor in the brain. This work describes our effort on the development of an albumin conjugate for enhanced inhibition effect on Aβ42 aggregation and amyloid cytotoxicity. Recently, several studies indicated that inhibitors with negative charges showed more effective suppressing effects on Aβ aggregation.7,32−34 For example, Liao et al.7 found that negatively charged gold NPs could serve as potential nanochaperones to inhibit and redirect Aβ fibrillization. Inspired by the findings, we herein propsose to modify bovine serum albumin (BSA) with carboxyl groups to increase the negative charges on the surface of the acidic protein (pI = 4.735,36). This led to the fabrication of acidulated BSA (A-BSA). The protein conjugate was extensively characterized and compared with native BSA to check the changes in molecular size, molecular structure, and stability. Then, the significantly enhanced inhibition effects of A-BSA on Aβ 42 aggregation and cytotoxicity were confirmed by extensive biophysical and biological assays.

z=

∑ Ni i

1 − 1 + 10 pH − pKi

∑ Mj j

10 pH − pKj 1 + 10 pH − pKj

(1)

where Ni and Mj are the numbers of basic and acidic groups, respectively, and pKi and pKj are the dissociation constants of basic and acidic groups, respectively. 2.3. Aβ42 Monomer Solution Preparation. The lyophilized Aβ42 protein was thawed at room temperature for 30 min, and then dissolved in HFIP to 1.0 mg/mL to remove pre-existing amyloid fibrils. The solution was then sonicated for 30 min followed by the removal of HFIP under vacuum for 24 h. The freeze-dried protein was immediately stored at −20 °C. The lyophilized protein was dissolved by 20 mM NaOH at 440 μM, sonicated for 20 min followed by centrifugation (16,000g) for 30 min at 4 °C. The upper 75% of the supernatant was carefully collected and then diluted 20-fold in buffer A to produce Aβ monomers, leading to a final protein concentration of 40 μM. This solution was used immediately for aggregation experiments. 2.4. Thioflavin T Fluorescent Assay. ThT is a benzothiazole dye that exhibits enhanced fluorescence (with excitation and emission at 440 and 480 nm, respectively) upon binding to amyloid fibrils and profibrils.39 The ThT fluorescence assay was conducted as described previously.21 Prior to the experiment, Aβ42 (40 μM) was mixed with BSA or A-BSA at different molar ratios. The mixture was incubated by continuous orbital shaking at 150 rpm and 37 °C. At different time points, 150 μL samples were drawn carefully, and 3 mL of ThT buffer (25 μM ThT in 25 mM sodium phosphate, pH 6.0 ± 0.05) was added into the cell and mixed uniformly. ThT fluorescence intensities were measured by a fluorescence spectrometer (Perking Elmer LS-55, MA, USA) with a slit width of 5 nm at 25 °C with excitation and emission at 440 and 480 nm, respectively. The measurements were performed in triplicate, and the data were averaged. The standard deviations were calculated and represented as error bars in figures. 2.5. Analysis by Size-Exclusion Chromatography. SEC analysis of Aβ42 aggregation was performed on an Ä KTA Explorer 100 system (GE Healthcare) using a Superdex 75 10/300 GL column. Aβ42 (40 μM) was incubated without or with A-BSA at a molar ratio of 1:1 by continuous orbital shaking at 150 rpm and 37 °C. Then, the solution was centrifuged at 16,000g for 10 min at 4 °C to remove insoluble aggregates. The supernatant was carefully drawn into a 1 mL syringe and then injected into the Superdex 75 10/300 GL column, using a 500-μL sample loop. Aβ42 species were eluted with buffer A at

2. MATERIALS AND METHODS 2.1. Materials. BSA, hexafluoroisopropanol (HFIP), phosphotungstic acid, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), diglycolic anhydride, thioflavin T (ThT), and dimethyl sulfoxide (DMSO) were purchased from Sigma (St. Louis, MO, USA). Aβ42 was purchased as lyophilized powder with a purity of >95% from GL Biochem (Shanghai, China). Sephadex G25 gel and Superdex 75 10/300 GL column were received from GE Healthcare (Uppsala, Sweden). The differentiated rat pheochromocytoma cells (PC12) were from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Dulbecco’s modified Eagle’s medium (DMEM) and fetal bovine serum (FBS) were obtained from Invitrogen (Carlsbad, CA, USA). Other chemicals were all of the highest purity available from local sources. 2.2. Synthesis and Characterization of Acidulated BSA. BSA was modified by reaction with diglycolic anhydride, as illustrated in Figure 1. The reaction conditions used in the literature37 in protein modification with anhydride were adopted for the modification. BSA solution (20 mL) was prepared by dissolving the protein powder in

Figure 1. Reaction scheme for the modification of BSA with diglycolic anhydride. 9790

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796

Langmuir

Article

a flow rate of 0.8 mL/min. The elution fractions were detected by UV absorbance at 215 nm at room temperature. The column was calibrated with BSA (67 kDa), green fluorescent protein (26.9 kDa), lysozyme (14.3 kDa), and vitamin B12 (1.36 kDa) under the same conditions. 2.6. Dynamic Light Scattering Measurements. Dynamic light scattering (DLS) measurements were performed on Zetasizer Nano (Malvern Instruments, Worchestershire, UK) to analyze the size distribution of Aβ42 aggregates. The samples for measurements were prepared by incubating 40 μM Aβ42 in the absence and presence of equimolar A-BSA in a shaking incubator at 150 rpm and 37 °C. The samples were submitted to size analysis by injecting 1.2 mL of the samples into disposable DLS cuvettes at 25 °C. The results were reported as the average of three measurements. 2.7. Analysis of Aβ42 Fibrils. During the incubation of 40 μM Aβ42 in the absence and presence of equimolar A-BSA (150 rpm, 37 °C), the optical density (OD) values of Aβ42 samples at 600 nm were measured by a UV/vis spectrometer (Perking Elmer Lambda 35, MA, USA) at 25 °C. The morphology of Aβ42 fibrils was observed by transmission electron microscopy (TEM). For a TEM observation, an Aβ42 sample (10 μL) was dropped on a carbon-coated copper grid (300-mesh) and air-dried for 5 min. The grid with amyloid fibrils was negatively stained with 2% (w/v) of phosphotungstic acid (pH 7.4) and air-dried. The samples were examined by a JEM100CXII transmission electron microscope system (JEOL Inc., Tokyo, Japan) at an accelerating voltage of 100 kV. 2.8. Cell Viability Assay. The MTT assay was employed to examine the cytotoxicity of Aβ42.40−42 The PC12 cells were incubated at 37 °C under 5% CO2 and were cultured in DMEM supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin in a CO2 cell culture box (NAPCO 5410, Tualatin, Oregon, USA). A total of 5 × 103 cells (90 μL) were seeded for 24 h in a polystyrene 96well plate. Then, the cells were treated with different concentrations of A-BSA or BSA, followed by freshly prepared Aβ42 stock solutions of a final protein concentration of 40 μM. The cells were incubated for an additional 24 h, and then a volume of 10 μL of MTT solution in 6 mg/ mL in phosphate buffered saline (PBS, 50 mM phosphate buffer containing 100 mM NaCl, pH 7.4) was added into each well and incubated for another 4 h. The medium was discarded, and 100 μL of DMSO was used to dissolve the cells until the purple crystals were fully dissolved. The absorbance at 570 nm was measured by a Plate Reader (TECAN Austria GmbH, Salzburg, Austria). The cell viability was calculated using the signals at 570 nm. Six replicates were performed, and the data were averaged. Background signals from sample treatment without cells were subtracted. The cell survival treated with PBS only was used as a basis of 100% to normalize other data for comparison. Analysis of variance followed by Student’s t-test was conducted for statistical comparisons, and p < 0.05 or less was considered to be statistically significant.

from 4.7 of native BSA. Under the physiological conditions (pH 7.4), however, the zeta potential changed less than those at lower pH values, which would be in favor of maintaining its native conformation. Although the conversion of amino groups into carboxyl groups did not result in the significant decrease of the surface zeta potential, considering the heterogeneous charge distributions on the protein surface, the local charge densities would have changed significantly by the modification. Then, the size distributions and intrinsic fluorescence of BSA and A-BSA were compared. At the starting point (0 h), BSA was 7.9 nm, while A-BSA was 8.1 nm (Figure S3A, Supporting Information). By incubating at 37 °C for 48 h, BSA became 8.5 nm, while A-BSA was 8.2 nm (Figure S3B, Supporting Information). The results indicate that the difference between the sizes of BSA and A-BSA was negligible and that A-BSA remained stable; little A-BSA aggregation occurred during the incubation. The intrinsic fluorescence spectra of BSA and A-BSA are shown in Figure S4 (Supporting Information). The fluorescence intensity and emission wavelength of BSA and A-BSA were almost the same at the beginning (0 h), and 48 h of incubation did not result in the changes of the intrinsic fluorescence spectra. The results indicate that the tertiary structure of A-BSA was consistent with that of native BSA and kept stable in the incubation. The above characterizations revealed the stability of A-BSA in both physical and chemical properties under physiological conditions. This ensures that A-BSA worka as a stable and biocompatible agent. 3.2. A-BSA Inhibits Aβ42 Fibrillogenesis. At first, Aβ42 solutions at a final concentration of 40 μM were coincubated with equimolar A-BSA or BSA, and Aβ42 fibrillations were monitored by ThT assay. As shown in Figure 2, Aβ42 showed a

Figure 2. Aggregation kinetics of Aβ42 incubated with and without BSA and A-BSA by ThT assays. The incubations were done in buffer A at 37 °C. Aβ42 concentration was 40 μM. Each data point was the mean of three different experiments, and the error bars represent standard deviations.

3. RESULTS AND DISCUSSION 3.1. Characteristics of A-BSA. We analyzed the modification degree of A-BSA by mass spectrometric analysis of BSA and A-BSA. As shown in Figure S1 (Supporting Information), a molecular weight increase of A-BSA over BSA by 1683 Da was observed due to the coupling of diglycolic anhydride (Figure 1). Because the molecular weight increase for coupling one diglycolic anhydride molecule was 116 Da, this molecular weight increase corresponds to an average modification degree of 14.5. In other words, in each BSA molecule, 14.5 amino groups were converted into carboxyl groups by the modification (Figure 1). The modification would lead to the increase of negative charges on A-BSA. To confirm this, the zeta potential values of BSA and A-BSA as a function of pH were measured (Figure S2, Supporting Information). As compared with BSA, A-BSA showed lower zeta potential values at the whole pH range (3.0−9.0), and the isoelectric point of A-BSA decreased to 4.0

fast growth phase, and a steady state was reached in about 24 h. In the presence of BSA or A-BSA, the growth rates of the ThT intensities decreased obviously, and the steady-state ThT levels also decreased significantly. Particularly, A-BSA showed stronger inhibitory potency than BSA; it reduced the ThT intensity by about 47%, while BSA only reduced by about 25%. The results indicate that the carboxyl groups on the A-BSA surface contributed to the occurrence of a stronger inhibition effect. Next, the effect of A-BSA concentration was investigated by the ThT assay (Figure 3). As can be seen, the inhibitory effect of A-BSA increased with its concentration. When its concentration reached 80 μM, the ThT intensity was reduced 9791

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796

Langmuir

Article

large Aβ42 aggregates (>70 kDa) because it was eluted in the void volume of the gel-filtration column (7.5 mL), which was packed with Superdex 75 gel with an exclusion limit of 70 kDa. The elution volumes of the two peaks are in good agreement with the previous study,46,47 in which the second peak was considered as monomeric Aβ42. It is seen from Figure 4 that, with increasing incubation time, more and more Aβ 42 aggregated into large oligomers (the first peak area increased) in the absence of A-BSA. However, more Aβ42 was in the monomeric state in the presence of A-BSA. This indicates that A-BSA effectively slowed down the aggregation rate of Aβ42 and maintained more Aβ42 in its monomeric state. To further confirm the inhibitory effect of A-BSA, the particle size distributions of Aβ42 aggregates were determined (Figure 5). Hydrodynamic diameters of the aggregate particles

Figure 3. Aggregation kinetics of Aβ42 incubated without and with ABSA at different concentrations: (□) Aβ only, (○) A-BSA/Aβ = 1:2, (Δ) A-BSA/Aβ = 1:1, and (▽) A-BSA/Aβ = 2:1. The incubations were performed in buffer A at 37 °C. Aβ42 concentration was 40 μM. Each data point was the mean of three different experiments, and the error bars represent standard deviations.

by about 57%. This indicates that the inhibition effect of A-BSA on Aβ42 fibrillation was concentration-dependent. This is similar to other chemical inhibitors, peptides, and NPs.7,43−45 The concentration-dependent manner was mainly reflected in the elongation phase of the fibrils. Therefore, it is considered that the increased negative charges on A-BSA were beneficial in the inhibition of Aβ aggregation in the elongation phase. Then, molecular weights of the soluble Aβ42 species were analyzed by using SEC. As shown in Figure 4, there were two elution peaks in the chromatogram. The first peak was relatively

Figure 5. DLS analysis of the Aβ species after incubation for (A) 24 h and (B) 48 h in the absence and presence of equimolar A-BSA (40 μM).

in solution were achieved via measuring the intensity of scattered lights. Although Aβ42 fibrils are heterogeneous and of long rod-like shapes,7 DLS analysis can provide a qualitative estimation of their size distribution. It is noted that the peak around 9 nm in Figure 5 was that of A-BSA. By 24 h of incubation (Figure 5A), a single peak at 400.7 nm was observed in the absence of A-BSA, while the size of aggregates was smaller in the presence of A-BSA, around 348.6 nm on average. After 48 h, the particle size of Aβ42 aggregation increased to 527.9 nm in the absence of A-BSA, while the presence of ABSA resulted in a smaller size, 399.8 nm (Figure 5B). The diminishment of larger aggregates in the presence of A-BSA reflected significant population changes in Aβ42 aggregation. It suggests that Aβ42 aggregation was greatly disturbed by A-BSA. Furthermore, the fibril morphology was observed by TEM (Figure 6). After incubation for 24 h, serried and entangled fibrils formed in the absence of A-BSA (Figure 6A), but Aβ42 fibrils became shorter and appeared in some small spherical, rod, and amorphous structures in the presence of A-BSA (Figure 6B). In addition, the fibrils stayed in a separated state from each other (Figure 6B). The results show that A-BSA

Figure 4. SEC analysis of soluble Aβ species after incubation for (A) 0 h, (B) 6 h, and (C) 12 h in the absence and presence of equimolar ABSA (40 μM). 9792

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796

Langmuir

Article

results indicate that A-BSA could prevent cell death caused by those detrimental Aβ42 species, and it is a potent biological inhibitor. 3.4. Mechanistic Model. To analyze the inhibition mechanism of A-BSA, the effect of NaCl concentration was investigated by using the ThT assay. It is obvious from Figure 8A that low salt concentration was favorable for the inhibition

Figure 6. TEM images of 40 μM Aβ42 incubated (A) without and (B) with 40 μM A-BSA for 24 h. Incubations were performed in buffer A at 37 °C.

changed the morphology of Aβ42 aggregates. This corresponds well with the SEC and DLS results discussed above. Therefore, it can be concluded from the results shown in Figures 2 to 6 that A-BSA efficiently inhibited the aggregation of Aβ42 and might have changed the aggregation pathway because of the morphological changes in the aggregates. These are expected to greatly reduce the cytotoxicity of Aβ42, which will be discussed in the next section. 3.3. A-BSA Significantly Mitigates Aβ42 Cytotoxicity. To assess the cytotoxicities of A-BSA and A-BSA induced Aβ42 aggregates, MTT assays were performed. As shown in Figure S5 (Supporting Information), the cell viability changed very little with increasing A-BSA concentration, indicating that ABSA is a biocompatible agent without cytotoxicity. By contrast, Aβ42 aggregates alone led to about 50% cell death (Figure 7).

Figure 8. Effect of NaCl concentration on the inhibition effect of ABSA. (A) (■) Aβ42 with 0 mM NaCl, (●) Aβ42 with 100 mM NaCl, (▲) Aβ42 with 200 mM NaCl, (□) A-BSA/Aβ42 = 1:1 with 0 mM NaCl, (○) A-BSA/Aβ42 = 1:1 with 100 mM NaCl, and (Δ) A-BSA/ Aβ42 = 1:1 with 200 mM NaCl. The incubations were done in 10 mM phosphate buffer (pH 7.4) with different NaCl concentrations at 37 °C. (B) Detailed data description of the figure before 6 h.

effect of A-BSA. When the salt concentration was increased to 200 mM, little inhibition effect remained. The result indicates that electrostatic interactions played a dominant role in the inhibition effects of A-BSA on Aβ42 fibrillogenesis. The process of Aβ42 fibrillogenesis can be divided into two phases, that is, nucleation phase and elongation phase.50 According to the results in Figure 8B, it is found that A-BSA displayed a good inhibition effect in the nucleation phase no matter what the NaCl concentration was at the first 4 h. It is considered that the binding of Aβ42 onto A-BSA was the main factor for the inhibition effect at this stage. Hydrophobic interactions contributed to the binding, so it was not compromised by increasing salt concentration. By increasing NaCl concentration, however, the inhibition effect of A-BSA gradually decreased in the elongation phase (about 4 to 24 h in Figure 8A). This clearly indicates that electrostatic interactions from the negative charges on A-BSA surface played a significant role in this elongation stage. To get a clearer picture about the effect of A-BSA, the turbidity changes of different solutions with incubation time were measured. Figure S6 (Supporting Information) shows that in the presence of A-BSA, there was a significant turbidity increase during the incubation of Aβ42. By analyzing the UV absorbance at 215 nm of the supernatants obtained after centrifugation, no distinct differences were observed between the incubations of A-BSA with and without Aβ42 (Table 1).

Figure 7. Viability of PC-12 cells incubated with 40 μM Aβ42 and different BSA and A-BSA concentrations. The incubations were performed in buffer A at 37 °C. ***, p < 0.001 as compared to the control group. ###, p < 0.001 as compared to Aβ42-treated group.

With increasing BSA and A-BSA concentration, the cell viability gradually increased. Particularly, A-BSA showed much stronger inhibitory potency than BSA. For example, as compared to the Aβ42-treated group, the cell viability increased by about 76% with equimolar A-BSA (40 μM), much higher than that (45%) with equimolar BSA. Moreover, the inhibition effect of A-BSA was much higher than that of other biological agents. For example, peptide inhibitors just caused 13% to 47% increases,48 and laminin only gave rise to about a 3% increase49 in cell viability of PC12 cells at the same molar ratio to Aβ42. The 9793

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796

Langmuir

Article

Table 1. A-BSA Concentration and Content in the Supernatants of the Incubations of A-BSA with and without Aβ42a incubation time (h)

A-BSA (mAU·mL)

A-BSA + Aβ42 (mAU·mL)

0 24 48

3565.7 ± 21.5 3514.1 ± 9.7 3497.5 ± 77.9

3591.3 ± 41.6 3503.5 ± 94.8 3492.3 ± 31.9

a

Albumin concentration is expressed by mAU at 215 nm, while content is represented by mAU·mL (mean ± SD, n = 3).

This indicates that it was almost all Aβ42 in the precipitates with little A-BSA coprecipitated. The result implies that A-BSA changed the pathway of Aβ42 aggregation, leading to the formation of off-pathway aggregates that precipitated during the incubation. This corresponds well with the TEM observation (Figure 6B). With the above results, a mechanistic model for the A-BSA effect on Aβ42 fibrillogenesis is proposed and represented in Figure 9. For a better understanding of the mechanistic model, the structures of Aβ42 and BSA were first analyzed. The amino acid sequence of Aβ42 is shown in Figure S7A (Supporting Information). The protein has two distinct hydrophobic parts, one in the central part (L17-S26) and the other on the C terminal (I31-A42).51 In aqueous solution, it displays a β-sheet structure before aggregating into the toxic oligomers and fibrils at physiological conditions (Figure S7B, Supporting Information).52 Surface analyses indicate that the two proteins have heterogeneous distributions of charges and hydrophobic patches (Figure S8, Supporting Information). From eq 1 and the pKa data given in Table S1 (Supporting Information), the net charge of BSA at pH 7.4 was calculated to be −13.4, while that of A-BSA was estimated at about −42 because 14.5 on average positively charged amino groups on BSA had been converted into carboxyl groups. Hence, A-BSA has much larger negatively charged areas than BSA as shown in Figure S8A (Supporting Information). Aβ42 has an isoelectric point of 5.5,53 so it also carries net negative charges at pH 7.4 (Figure S8B, Supporting Information), and the net charge number was calculated from eq 1 to be −3.2 at pH 7.4. Therefore, Aβ42 and A-BSA or Aβ42 and BSA are electrostatically repulsed from each other. However, hydrophobic interactions at specific areas can also lead to the binding of Aβ42 onto A-BSA or BSA. Namely, there exist two opposite forces between bound Aβ42 and A-BSA or bound Aβ42 and BSA at physiological conditions. This is the basis for the development of the mechanistic model for BSA and A-BSA effects on Aβ42 aggregation (Figure 9B and C). It has been recognized that in on-pathway fibrillogenesis, Aβ42 monomers assemble into amyloid fibrils via several metastable oligomers (Figure 9A).7 In the presence of BSA (Figure 9B), hydrophobic binding of Aβ42 onto the protein would reduce free Aβ42 monomer concentration to some extent in the nucleation phase,54 thereby reducing the nucleation rate, as shown in Figure 2. However, when Aβ42 was bound on the protein surface, electrostatic repulsion between the two proteins would give rise to the conformational changes of Aβ42, which might affect the aggregation process to some extent. However, because of the wide distribution of positively charged patches in the vicinity of hydrophobic patches (Figure S8A, Supporting Information), the local electrostatic repulsion between BSA and Aβ42 would be distributed not widely enough to cause significant changes of Aβ42 conformation and the aggregation process, so the electrostatic effect of BSA on Aβ42

Figure 9. Schematic representations of Aβ42 fibrillation pathways. (A) On-pathway aggregation: Aβ42 monomers assemble into fibrils and cause neurotoxicity. (B) Aβ42 aggregation in the presence of BSA: BSA binding reduces the nucleation rate but has less influence on the elongation phase. (C) Aβ42 aggregation in the presence of A-BSA: hydrophobic binding and electrostatic repulsion by A-BSA (see D below) lead to the conformational changes of Aβ42 and make the βsheet conformation unstable, slowing down the aggregation and leading to the off-pathway aggregation. (D) Schematic representation of the hydrophobic binding of Aβ42 at the hydrophobic regions and electrostatic repulsion. The opposite forces make Aβ42 monomers stay in extended conformations.

aggregation was not significant. Namely, the inhibition effect of native BSA on Aβ aggregation is mainly due to its binding of Aβ (Figure 9B). In the presence of A-BSA (Figure 9C), however, more widely distributed negative charges would give rise to broad local repulsions for the bound Aβ42, and the attractive and repulsive forces become well-balanced, resulting in remarkable changes of Aβ42 conformation on the A-BSA surface (Figure 9D). The hydrophobic binding of the two hydrophobic regions on Aβ42 and the electrostatic repulsion are illustrated in Figure 9D. The two well-balanced forces make Aβ42 display extended conformations, which are distinctly different from the β-sheet structure (Figure S7B, Supporting Information) that is necessary for the on-pathway fibrillogenesis in Figure 9A.55 The extended structure can be maintained even when Aβ42 is dissociated into the bulk solution. Hence, A-BSA can greatly slow down Aβ42 aggregation and change the fibrillogenesis pathway into the formation of small spherical, rod, and amorphous structures (Figure 6B) that have less cytotoxicity (Figure 7). 9794

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796

Langmuir

Article

(2) Vitiello, G.; Di Marino, S.; D’Ursi, A. M.; D’Errico, G. Omega-3 fatty acids regulate the interaction of the Alzheimer’s abeta(25-35) peptide with lipid membranes. Langmuir 2013, 29, 14239−14245. (3) Jakob-Roetne, R.; Jacobsen, H. Alzheimer’s disease: from pathology to therapeutic approaches. Angew. Chem., Int. Ed. 2009, 48, 3030−3059. (4) Sehgal, N.; Gupta, A.; Valli, R. K.; Joshi, S. D.; Mills, J. T.; Hamel, E.; Khanna, P.; Jain, S. C.; Thakur, S. S.; Ravindranath, V. Withania somnifera reverses Alzheimer’s disease pathology by enhancing lowdensity lipoprotein receptor-related protein in liver. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 6. (5) Brambilla, D. PEGylated nanoparticles bind to and alter amyloidbeta peptide conformation: toward engineering of functional nanomedicines for Alzheimer’s disease. ACS Nano 2012, 6, 5897−5908. (6) Thapa, A.; Vernon, B. C.; De la Pena, K.; Soliz, G.; Moreno, H. A.; Lopez, G. P.; Chi, E. Y. Membrane-mediated neuroprotection by curcumin from amyloid-beta-peptide-induced toxicity. Langmuir 2013, 29, 11713−11723. (7) Liao, Y. H.; Chang, Y. J.; Yoshiike, Y.; Chang, Y. C.; Chen, Y. R. Negatively charged gold nanoparticles inhibit Alzheimer’s amyloidbeta fibrillization, induce fibril dissociation, and mitigate neurotoxicity. Small 2012, 8, 3631−3639. (8) Tanzi, R. E.; Bertram, L. Twenty years of the Alzheimer’s disease amyloid hypothesis: a genetic perspective. Cell 2005, 120, 545−555. (9) Richman, M.; Perelman, A.; Gertler, A.; Rahimipour, S. Effective targeting of Abeta to macrophages by sonochemically prepared surface-modified protein microspheres. Biomacromolecules 2013, 14, 110−116. (10) Jonsson, T.; Atwal, J. K.; Steinberg, S.; Snaedal, J.; Jonsson, P. V.; Bjornsson, S.; Stefansson, H.; Sulem, P.; Gudbjartsson, D.; Maloney, J.; Hoyte, K.; Gustafson, A.; Liu, Y.; Lu, Y.; Bhangale, T.; Graham, R. R.; Huttenlocher, J.; Bjornsdottir, G.; Andreassen, O. A.; Jonsson, E. G.; Palotie, A.; Behrens, T. W.; Magnusson, O. T.; Kong, A.; Thorsteinsdottir, U.; Watts, R. J.; Stefansson, K. A mutation in APP protects against Alzheimer’s disease and age-related cognitive decline. Nature 2012, 488, 96−99. (11) Jiang, D.; Rauda, I.; Han, S.; Chen, S.; Zhou, F. Aggregation pathways of the amyloid beta(1-42) peptide depend on its colloidal stability and ordered beta-sheet stacking. Langmuir 2012, 28, 12711− 12721. (12) Zhang, Y. W.; Thompson, R.; Zhang, H.; Xu, H. APP processing in Alzheimer’s disease. Mol. Brain 2011, 4, 3. (13) Yu, X.; Wang, Q.; Lin, Y.; Zhao, J.; Zhao, C.; Zheng, J. Structure, orientation, and surface interaction of Alzheimer amyloid-beta peptides on the graphite. Langmuir 2012, 28, 6595−6605. (14) Stanyon, H. F.; Viles, J. H. Human Serum albumin can regulate amyloid-beta peptide fiber growth in the brain interstitium: implications for Alzheimer disease. J. Biol. Chem. 2012, 287, 28163− 28168. (15) Cramer, P. E.; Cirrito, J. R.; Wesson, D. W.; Lee, C. Y.; Karlo, J. C.; Zinn, A. E.; Casali, B. T.; Restivo, J. L.; Goebel, W. D.; James, M. J.; Brunden, K. R.; Wilson, D. A.; Landreth, G. E. ApoE-directed therapeutics rapidly clear beta-amyloid and reverse deficits in AD mouse models. Science 2012, 335, 1503−1506. (16) Yoo, S. I.; Yang, M.; Brender, J. R.; Subramanian, V.; Sun, K.; Joo, N. E.; Jeong, S. H.; Ramamoorthy, A.; Kotov, N. A. Inhibition of amyloid peptide fibrillation by inorganic nanoparticles: functional similarities with proteins. Angew. Chem., Int. Ed. 2011, 50, 5110−5115. (17) Adessi, C.; Frossard, M. J.; Boissard, C.; Fraga, S.; Bieler, S.; Ruckle, T.; Vilbois, F.; Robinson, S. M.; Mutter, M.; Banks, W. A.; Soto, C. Pharmacological profiles of peptide drug candidates for the treatment of Alzheimer’s disease. J. Biol. Chem. 2003, 278, 13905− 13911. (18) Sciarretta, K. L.; Gordon, D. J.; Meredith, S. C. Peptide-based inhibitors of amyloid assembly. Methods Enzymol. 2006, 413, 273−312. (19) Palhano, F. L.; Lee, J.; Grimster, N. P.; Kelly, J. W. Toward the molecular mechanism(s) by which EGCG treatment remodels mature amyloid fibrils. J. Am. Chem. Soc. 2013, 135, 7503−7510.

Protein unfolding caused by simultaneous hydrophobic binding and electrostatic repulsion in opposite directions has been observed in molecular simulation studies on hydrophobic charge induction chromatography by using a coarse-grained model protein.56,57 When the hydrophobically bound protein was dissociated from the surface by electrostatic repulsion, protein unfolding occurred and the unfolded conformation was still observed in the vicinity of the surface due to the long-range nature of electrostatic interaction. The computational studies offered strong theoretical support on the mechanistic model we proposed herein.

4. CONCLUSIONS We have developed acidulated serum albumin as a biological inhibitor of Aβ42 fibrillogenesis. A-BSA carries much more negative charges than BSA and maintains the stability and biocompatibility as BSA does. It was found that A-BSA could slow down aggregation and change the fibrillation pathway of Aβ42, leading to the formation of less toxic aggregates. Comparisons to literature data suggest that A-BSA was more effective than other biological agents in the inhibition of Aβ42 aggregation. On the basis of extensive biophysical and biological analyses, a mechanistic model for describing the inhibition effect of A-BSA on Aβ42 aggregation and cytotoxicity was developed. The model considers that there exist two opposite forces for Aβ42 bound on the A-BSA surface, that is, hydrophobic binding and electrostatic repulsion, due to the wide distributions of the hydrophobic patches and negatively charged areas on A-BSA. The well-balanced forces give rise to the extension of Aβ42 conformations, making Aβ42 distinctly different from the β-sheet structure that is necessary for onpathway fibrillogenesis. Thus, A-BSA works as a potent biomacromolecular inhibitor of Aβ42 fibrillation. The findings and the mechanistic model would benefit in the design of more potent inhibitors of amyloid β-protein fibrillogenesis and cytotoxicity.

■

ASSOCIATED CONTENT

S Supporting Information *

Mass spectra, zeta potentials, size distributions, and fluorescence intensities of BSA and A-BSA, cytotoxicity assay of ABSA, turbidity analysis of Aβ42 samples, amino acid sequence and β-sheet structure of Aβ42, and surface models of BSA and Aβ42. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +86 22 27404981. Fax: +86 22 27403389. E-mail: ysun@ tju.edu.cn. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was funded by the Natural Science Foundation of China (Nos. 21236005 and 21376172).

■

REFERENCES

(1) Selkoe, D. J. Alzheimer’s disease: genes, proteins, and therapy. Physiol. Rev. 2001, 81, 26. 9795

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796

Langmuir

Article

(40) Shearman, M. S. Toxicity of protein aggregates in PC12 cells: 3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay. Methods Enzymol. 1999, 309, 716−723. (41) Zhang, C.; Wan, X.; Zheng, X.; Shao, X.; Liu, Q.; Zhang, Q.; Qian, Y. Dual-functional nanoparticles targeting amyloid plaques in the brains of Alzheimer’s disease mice. Biomaterials 2014, 35, 456−465. (42) Chen, Q.; Yang, L.; Zheng, C.; Zheng, W.; Zhang, J.; Zhou, Y.; Liu, J. Mo polyoxometalate nanoclusters capable of inhibiting the aggregation of Abeta-peptide associated with Alzheimer’s disease. Nanoscale 2014, 6, 6886−6897. (43) Veloso, A. J.; Kerman, K. Modulation of fibril formation by a beta-sheet breaker peptide ligand: an electrochemical approach. Bioelectrochemistry 2012, 84, 49−52. (44) Giordano, C.; Sansone, A.; Masi, A.; Masci, A.; Mosca, L.; Chiaraluce, R.; Pasquo, A.; Consalvi, V. Inhibition of amyloid peptide fragment Aβ25−35 fibrillogenesis and toxicity by N-terminal β-amino acid-containing esapeptides: is taurine moiety essential for in vivo effects? Chem. Biol. Drug. Des. 2012, 79, 30−37. (45) Zhang, W.; Du, Y.; Bai, M.; Xi, Y.; Li, Z.; Miao, J. S14Ghumanin inhibits Abeta1-42 fibril formation, disaggregates preformed fibrils, and protects against Abeta-induced cytotoxicity in vitro. J. Pept. Sci. 2013, 19, 159−165. (46) Jan, A.; Hartley, D. M.; Lashuel, H. A. Preparation and characterization of toxic Abeta aggregates for structural and functional studies in Alzheimer’s disease research. Nat. Protoc. 2010, 5, 1186− 1209. (47) Kim, T.; Vidal, G. S.; Djurisic, M.; William, C. M.; Birnbaum, M. E.; Garcia, K. C.; Hyman, B. T.; Shatz, C. J. Human LilrB2 is a betaamyloid receptor and its murine homolog PirB regulates synaptic plasticity in an Alzheimer’s model. Science 2013, 341, 1399−1404. (48) Li, H.; Du, Z.; Lopes, D. H.; Fradinger, E. A.; Wang, C.; Bitan, G. C-terminal tetrapeptides inhibit Abeta42-induced neurotoxicity primarily through specific interaction at the N-terminus of Abeta42. J. Med. Chem. 2011, 54, 8451−8460. (49) Monji, A.; Tashiro, K.-i.; Yoshida, I.; Kaname, H.; Hayashi, Y.; Matsuda, K.; Tashiro, N. Laminin inhibits both Aβ40 and Aβ42 fibril formation but does not affect Aβ40 or Aβ42-induced cytotoxicity in PC12 cells. Neurosci. Lett. 1999, 266, 85−88. (50) Cabaleiro-Lago, C.; Quinlan-Pluck, F.; Lynch, I.; Lindman, S.; Minogue, A. M.; Thulin, E.; Walsh, D. M.; Dawson, K. A.; Linse, S. Inhibition of amyloid beta protein fibrillation by polymeric nanoparticles. J. Am. Chem. Soc. 2008, 130, 15437−15443. (51) Luhrs, T.; Ritter, C.; Adrian, M.; Riek-Loher, D.; Bohrmann, B.; Dobeli, H.; Schubert, D.; Riek, R. 3D structure of Alzheimer’s amyloidbeta(1-42) fibrils. P. Natl. Acad. Sci. U.S.A. 2005, 102, 17342−17347. (52) Pellarin, R.; Caflisch, A. Interpreting the Aggregation Kinetics of Amyloid Peptides. J. Mol. Biol. 2006, 360, 882−892. (53) Wood, S. J.; Maleeff, B.; Hart, T.; Wetzel, R. Physical, morphological and functional differences between pH 5.8 and 7.4 aggregates of the Alzheimer’s amyloid peptide Aβ. J. Mol. Biol. 1996, 256, 870−877. (54) Algamal, M.; Milojevic, J.; Jafari, N.; Zhang, W.; Melacini, G. Mapping the interactions between the Alzheimer’s Abeta-peptide and human serum albumin beyond domain resolution. Biophys. J. 2013, 105, 1700−1709. (55) Yang, C.; Li, J. Y.; Li, Y.; Zhu, X. L. The effect of solvents on the conformations of amyloid beta-peptide (1-42) studied by molecular dynamics simulation. J. Mol. Struct. 2009, 895, 1−8. (56) Zhang, L.; Zhao, G.; Sun, Y. Effects of ligand density on hydrophobic charge induction chromatography: molecular dynamics simulation. J. Phys. Chem. B 2010, 114, 2203−2211. (57) Zhang, L.; Zhao, G.; Sun, Y. Molecular insight into protein conformational transition in hydrophobic charge induction chromatography: a molecular dynamics simulation. J. Phys. Chem. B 2009, 113, 6873−6880.

(20) Necula, M.; Kayed, R.; Milton, S.; Glabe, C. G. Small molecule inhibitors of aggregation indicate that amyloid beta oligomerization and fibrillization pathways are independent and distinct. J. Biol. Chem. 2007, 282, 10311−10324. (21) Chan, H. M.; Xiao, L.; Yeung, K. M.; Ho, S. L.; Zhao, D.; Chan, W. H.; Li, H. W. Effect of surface-functionalized nanoparticles on the elongation phase of beta-amyloid (1-40) fibrillogenesis. Biomaterials 2012, 33, 4443−4450. (22) Xiao, L.; Zhao, D.; Chan, W. H.; Choi, M. M.; Li, H. W. Inhibition of beta 1-40 amyloid fibrillation with N-acetyl-L-cysteine capped quantum dots. Biomaterials 2010, 31, 91−98. (23) Kerr, M. L.; Gasperini, R.; Gibbs, M. E.; Hou, X.; Shepherd, C. E.; Strickland, D. K.; Foa, L.; Lawen, A.; Small, D. H. Inhibition of Abeta aggregation and neurotoxicity by the 39-kDa receptor-associated protein. J. Neurochem. 2010, 112, 1199−1209. (24) Milojevic, J.; Raditsis, A.; Melacini, G. Human serum albumin inhibits Abeta fibrillization through a “monomer-competitor” mechanism. Biophys. J. 2009, 97, 2585−2594. (25) Palivan, C. G.; Fischer-Onaca, O.; Delcea, M.; Itel, F.; Meier, W. Protein-polymer nanoreactors for medical applications. Chem. Soc. Rev. 2012, 41, 2800−2823. (26) Klonoff-Cohen, H.; Barrett-Connor, E. L.; Edelstein, S. L. Albumin levels as a predictor of mortality in the healthy elderly. J. Clin. Epidemiol. 1992, 45, 207−212. (27) Inhee, C.; Lee, P.; Rapid, L. Detection of Aβ aggregation and inhibition by dual functions of gold nanoplasmic particles: catalytic activator and optical reporter. ACS Nano 2013, 7, 10. (28) Kuo, Y. M.; Kokjohn, T. A.; Kalback, W.; Luehrs, D.; Galasko, D. R.; Chevallier, N.; Koo, E. H.; Emmerling, M. R.; Roher, A. E. Amyloid-beta peptides interact with plasma proteins and erythrocytes: implications for their quantitation in plasma. Biochem. Biophys. Res. Commun. 2000, 268, 750−756. (29) Rozga, M.; Kloniecki, M.; Jablonowska, A.; Dadlez, M.; Bal, W. The binding constant for amyloid Abeta40 peptide interaction with human serum albumin. Biochem. Biophys. Res. Commun. 2007, 364, 714−718. (30) Elovaara, I.; Palo, J.; Erkinjuntti, T.; Sulkava, R. Serum and cerebrospinal fluid proteins and the blood-brain barrier in Alzheimer’s disease and multi-infarct dementia. Eur. J. Neurol. 1987, 26, 229−234. (31) Kay, A. D.; May, C.; Papadopoulos, N. M.; Costello, R.; Atack, J. R.; Luxenberg, J. S.; Cutler, N. R.; Rapoport, S. I. CSF and serum concentrations of albumin and IgG in Alzheimer’s disease. Neurobiol. Aging 1987, 8, 21−25. (32) Mahmoudi, M.; Quinlan-Pluck, F.; Monopoli, M. P.; Sheibani, S.; Vali, H.; Dawson, K. A.; Lynch, I. Influence of the physiochemical properties of superparamagnetic iron oxide nanoparticles on amyloid beta protein fibrillation in solution. ACS Chem. Neurosci. 2013, 4, 475− 485. (33) Kim, S. T.; Saha, K.; Kim, C.; Rotello, V. M. The role of surface functionality in determining nanoparticle cytotoxicity. Acc. Chem. Res. 2013, 46, 681−691. (34) Santa-María, I.; Hernández, F.; Moreno, F. J.; Avila, J. Taurine, an inducer for tau polymerization and a weak inhibitor for amyloid-βpeptide aggregation. Neurosci. Lett. 2007, 429, 91−94. (35) Xu, Y.; Takai, M.; Ishihara, K. Protein adsorption and cell adhesion on cationic, neutral, and anionic 2-methacryloyloxyethyl phosphorylcholine copolymer surfaces. Biomaterials 2009, 30, 4930− 4938. (36) Gu, M.; Kilduff, J. E.; Belfort, G. High throughput atmospheric pressure plasma-induced graft polymerization for identifying proteinresistant surfaces. Biomaterials 2012, 33, 1261−1270. (37) D.Powers, J. Trypsin purification by affinity binding to small unilamellar liposomes. Biotechnol. Bioeng. 1990, 36, 14. (38) Bujacz, A. Structures of bovine, equine and leporine serum albumin. Acta Crystallogr., Sect. D 2012, 68, 1278−12789. (39) Khurana, R.; Coleman, C.; Ionescu-Zanetti, C.; Carter, S. 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. 9796

dx.doi.org/10.1021/la5025197 | Langmuir 2014, 30, 9789−9796