Abstract
Accumulation of amyloid fibrils is one of the likely key factors leading to the development of Alzheimer's disease and other amyloidosis associated diseases. Magnetic nanoparticles (NPs) have been developed as promising medical materials for many medical applications. In this study, we have explored the effects of Fe3O4 NPs on the fibrillogenesis process of insulin fibrils. When Fe3O4 NPs were co-incubated with insulin, Fe3O4 NPs had no effect on the structural transformation into amyloid-like fibrils but had higher affinity toward insulin fibrils. We demonstrated that the zeta potential of insulin fibrils and Fe3O4 NPs were both positive, suggesting the binding forces between Fe3O4 NPs and insulin fibrils were van der Waals forces but not surface charge. Moreover, a different amount of Fe3O4 NPs added had no effect on secondary structural changes of insulin fibrils. These results propose the potential use of Fe3O4 NPs as therapeutic agents against diseases related to protein aggregation or contrast agents for magnetic resonance imaging.
Export citation and abstract BibTeX RIS
1. Introduction
Amyloid protein misfolding to form highly ordered fibrillary aggregates is believed to be the main characteristic feature of various neuropathological diseases, including Alzheimer's, Parkinson's, and Huntington's disease and a factor in other amyloidosis associated diseases, such as type II diabetes mellitus [1, 2]. It is well-known that self-assembly of beta amyloid (Aβ) peptide/protein into a cross β-sheeted fibril is associated with amyloid-related disorders [3, 4]. Understanding insulin fibrillation is essential for formulating insulin therapies. Recent work by Muzaffar et al reported that NaCl induced insulin fibrillation mainly results from subtle structural changes and is not a mere salt effect [5]. However, preventing fibrillogenesis as well as inhibiting the Aβ fibril aggregation has been considered as one therapeutic approach for these diseases.
Various potential approaches have been reported to decrease or eliminate amyloid fibrillation processes, including small molecules, functional polymers, and nanoparticles (NPs) [6–9]. The NPs' influence on protein fibrillation is a function of both the NP surface interfacial properties, including charge and its enormous surface area/volume [10, 11]. Wu et al demonstrated that only TiO2 NPs significantly enhance the rate of Aβ fibrillation but SiO2, ZrO2, CeO2, C60, and C70 have almost no effect on Aβ42 fibrillation [12]. Linse et al show that NPs (copolymer particles, cerium oxide particles, quantum dots, and carbon nanotubes) enhance the fibrillation of human β2-microglobulin [13]. In these cases, the NPs can act as catalysts for protein amyloid assembly, increasing the risk of amyloid diseases. Although NPs have been used to solve the problem of amyloid protein aggregation, the biological safety of NPs is not fully elucidated. One of the main factors in the toxicity of NPs to cells could be the size [14], concentration [15], and surface characteristics [16, 17]. Additionally, the adsorption ability of NPs may induce protein aggregation, which may be an important factor for the estimation of cytotoxicity to cells [18]. Magnetic NPs have been intensively developed, not only for their fundamental scientific interest, but also for many medical applications, such as magnetic resonance imaging, hyperthermia for tumor treatment, cell labeling and sorting, DNA separation and drug delivery [19]. However, so far only a few studies have been related to magnetic NPs and amyloid protein aggregation. Skaat et al showed that gelatin modified γ-Fe2O3 NPs are attached to the insulin fibrils leading to specific magnetization of fibrils, allowing the removal of the insulin amyloid fibrils from their continuous phase by magnetization, but did not show whether the interaction of the NPs with insulin fibers cause the conformational change or the formation of amyloid protein aggregates [20]. Furthermore, Bellova et al demonstrated that magnetic Fe3O4 NPs are able to inhibit the formation of lysozyme amyloid aggregates and have the potential to destroy amyloid fibrils by depolymerization of amyloid structures [21]. Iron oxide NPs are lowly toxic when compared with other metal oxide NPs, whereas, for example, Fe3O4 NPs are much less toxic than Fe2O3 NPs [22]. Furthermore, with the advantages of high magnetization, Fe3O4 NPs are widely used for magnetic separation [23]. However, the use of iron oxide NPs in amyloid protein aggregation is still extremely limited.
The aim of this study was to investigate the effect of magnetic NPs, Fe3O4 on the interaction of insulin amyloid fibrils.
2. Experimental methods
2.1. Preparation of Fe3O4 NPs
Fe3O4 NPs were prepared by a chemical precipitation method [19]. Stock solutions of 1 M FeCl3 · 6H2O and 0.5 M FeCl2 · 4H2O were prepared separately by dissolving the iron salts in 200 ml of distilled water at 80 °C. The solutions were then mixed and stirred at 265 rpm, while slowly adding 2 M NaOH until a pH of 14 was reached. The reaction proceeded for 1 h at 80 °C under a N2 atmosphere, and the resulting Fe3O4 NPs were purified repeatedly by magnetic field separation with excess distilled water until a neutral pH level was reached.
2.2. Fe3O4 NPs characterization
NP crystal structure characterization was carried out by powder x-ray diffraction (XRD) on a Bruker D8 Advance x-ray diffractometer. Transmission electron microscopy (TEM) images were acquired using a Hitachi HF-2000 TEM to determine particle size. Room-temperature magnetization curves for Fe3O4 NPs were acquired using a Quantum Design MPMS-XL7 superconducting quantum interference device. Saturation magnetization (Ms) for Fe3O4 NPs was approximately 43.2 emu g-1. X-ray photoelectron spectroscopy was performed on both types of NPs to determine the chemical state of Fe using a JEOL JAMP-9500F.
2.3. Insulin fibrils formation
Insulin powder from a bovine pancreas (Sigma–Aldrich, St.Louis, MO, USA) was used without further purification to prepare insulin fibrils. A stock 1 mg ml−1 insulin solution with (or without) Fe3O4 NPs was dissolved in HCl solution at pH 1.6 to make a 1000 ppm insulin/Fe3O4 (insulin) solution. The insulin solution was then incubated at 80 °C for 2.5 h using the method described by Jansen et al [24]. After incubation, the solutions were stored at room temperature for periods of 0–6 weeks. Samples were characterized at various room temperature incubation times using cryogenic TEM (Cryo-TEM, JEM-1400) and a circular dichroism (CD) spectrometer (CD, Jasco, J-810).
2.4. CD and Fourier transform infrared spectroscopy
To explore the secondary structure of insulin fibrils after interaction with Fe3O4 NPs, the CD spectra were measured with a 0.5 mm path length cell from 260–190 nm at 25 °C, and the data were averaged over three measurements.
Fourier transform infrared spectroscopy (FTIR) was also employed. The FTIR spectra were recorded on VERTEX 80/80v FTIR spectrometers (Bruker). For each spectrum, a 512-scan interferogram was collected in single-beam mode at 25 °C in a vacuum environment using a 4 cm−1 resolution.
2.5. Zeta potential analyzer
To examine the zeta potential of insulin fibrils and Fe3O4, a zeta potential analyzer was employed. The zeta potential was measured at 25 °C with a DelsaTM NanoC particle analyzer (BeckMan Coulter).
3. Results and discussion
In our study, insulin was chosen as a model of amyloidogenic protein to investigate whether Fe3O4 NPs affect insulin fibrils aggregation. We exclude factors for aggregating formation by simplification of insulin solution, and the concentration of Fe3O4 NPs was used at 0.1 ∼ 2.0 μg ml−1, which is far less than the toxicity of 100 μg ml−1 [25]. Fe3O4 NPs were synthesized by using a co-precipitation method, and coating without any material.
First, the characterization of Fe3O4 NPs was determined. Fe3O4 NPs samples were prepared by a chemical precipitation method [26], and XRD spectra were employed to determine their crystal structure. As shown in figure 1(a), the crystallized structure of Fe3O4 spectra was 2θ: 30.1°, 35.4°, 43.1°, 53.4°, 57°, 62.6°, and 75.3°, which are assigned to the (220), (311), (400), (422), (511), (440), and (533) crystallographic faces, respectively (figure 1(a)). These designations and values are in good accordance with the standard card (JCPDS card no. 19-0629). The size of the Fe3O4 NPs was determined by TEM. TEM images showed that the average size of Fe3O4 NPs was 14.9 ± 2.4 nm (figure 1(b)), which is a suitable size for medical application.
Figure 1. XRD spectra of Fe3O4 (a) and TEM images of Fe3O4 (b).
Download figure:
Standard image High-resolution imageIn order to determine the effect of Fe3O4 NPs on the fibrillation process, we studied insulin fibrils self-assembled into amyloid-like fibrils. Samples were prepared by incubating insulin monomers at pH1.6 and 80 °C for 2.5 h, which were then stored at room temperature for periods of 0–6 weeks to monitor the conformational change of insulin fibrils and the affinity of Fe3O4 NPs for insulin fibrils. Figures 2(a) and (b) show the CD spectra from insulin samples prepared without (a) and with (b) Fe3O4. There was no significant difference between the spectra for either pre-treated samples, so it was suggested that Fe3O4 NPs did not affect the incubation period of insulin conformation transition. The results from the CD spectra indicated that the incubation period of the structural transformation from the α-helix to the β-sheet conformation required about 5–6 weeks.
Figure 2. CD spectra of insulin fibrillation in the absence (a) or presence (b) of Fe3O4 after incubation for 2.5 h, 1–6 weeks at pH 1.6 and 80 °C.
Download figure:
Standard image High-resolution imageTo further confirm whether Fe3O4 NPs bound to insulin fibrils during the conformational change, TEM was used to characterize the morphological variation of the aggregates. As shown in figure 3, we observed that Fe3O4 NPs bound to insulin fibrils after 2.5 h incubation. TEM images revealed the interesting phenomena that Fe3O4 NPs bound to either the α-helix or the β-sheet of insulin fibrils (figures 3(b)∼(g)). Moreover, TEM images showed that Fe3O4 NPs were stable during the fibrillation process and no aggregates occurred (figures 3(b)∼(g)).
Figure 3. TEM images of Fe3O4 bound to insulin fibrils incubation in a solution at pH 1.6 and 80 °C for 2.5 h, and then stored at room temperature for (a) 0, (b) 1, (c) 2, (d) 3, (e) 4, (f) 5, and (g) 6 weeks. The lower left panel shows a diagram of Fe3O4 absorbed on fibrils. The scale bar in each image is 200 nm.
Download figure:
Standard image High-resolution imageFurthermore, we used magnetic separation techniques to prove that Fe3O4 NPs had a higher affinity with insulin fibrils. As shown in figures 4(a) and (c), magnetic Fe3O4 NPs were removed from an aqueous solution by using magnets. In order to further confirm the interaction of Fe3O4 NPs with insulin fibrils, we stained the insulin fibrils with Congo Red, a histologic dye that binds to β-sheet insulin fibrils [27]. As shown in figure 4(b), a blue-purple color is displayed in an aqueous solution of pH 1.6. Figure 4(d) indicates that blue-purple insulin fibrils bound with Fe3O4 NPs were attracted by magnets, suggesting Fe3O4 NPs had a higher affinity with insulin fibrils.
Figure 4. (a) Incubated with an aqueous solution of bovine insulin and Fe3O4 NPs, (b) stained insulin fibrils with Congo Red, (c) (d) magnetic separation of the Fe3O4 from an aqueous solution.
Download figure:
Standard image High-resolution imageThe types of binding force acting between the NPs and protein fibers include electrostatic interaction force or van der Waals forces [12]. To better understand the types of force involved in the interaction of Fe3O4 NPs with insulin fibrils, we first determined whether surface charge affected the interaction of Fe3O4 NPs with insulin fibrils. Znidarsic et al found that the zeta potential is positive due to the fact that the N-terminal residue of protein is –NH3 when protein appears under its isoelectric point (IEP) [28]. The zeta potential of Fe3O4 NPs at pH > 6.3 is negative, while that of Fe3O4 NPs at pH < 6.3 is positive [29]. A series of experiments was performed in solution at pH = 1.6 in this study, and the IEP of bovine insulin was pH = 5.323 [30]. Previous studies demonstrated that the zeta potentials of bovine insulin and Fe3O4 NPs were positive in solution at pH = 1.6. Our results showed that the zeta potentials of insulin fibrils was +3.44 mV, while the zeta potentials of Fe3O4 NPs was +25.7 mV (table 1), suggesting that the binding force acting between the NPs and protein fibers was not due to surface charges, but caused by van der Waals forces.
Table 1. The zeta potentials of Fe3O4 NPs and insulin fibrils in an aqueous solution.
| Material | Solvent | Zeta Potential (mV) |
|---|---|---|
| Insulin (α) | HCl (pH 1.6) | 3.44 |
| Insulin (β) | HCl (pH 1.6) | 17.30 |
| Fe3O4 | H2O | −9.76 |
| HCl (pH 1.6) | 25.70 | |
| +Insulin (α) | 25.24 | |
| +Insulin (β) | 29.30 |
In order to assess the effects of Fe3O4 NPs on an increase in the interaction of Fe3O4 NPs with insulin fibrils or any secondary structural changes of insulin fibrils, we used Fe3O4 NPs concentrations of 200, 400, 600, 800 and 1000 ppm in an aqueous solution before the fibrillation process, and then incubated at 80 °C for 2.5 h. The CD spectra in figure 5(a) revealed that there was no significant difference between the spectra for the different amounts added. Moreover, the different amounts of Fe3O4 NPs added had no effect on the secondary structural changes of insulin fibrils, which still displayed α-helix. Additionally, in the absorption spectra shown in figure 5(b), Fe3O4 NPs bound to insulin fibrils and exhibited two peaks at 630 cm−1 and 581 cm−1, which were Fe3O4 tetrahedral and Fe3O4 octahedral [30]. The peaks between 1220 and 1170 cm−1 were assigned to the C–N and C–O stretching bands of protein, respectively. The peak at 1450 cm−1 corresponded to the CH2 bending band of protein. The amide I band and the amide II band were assigned to 1650 cm−1 and 1542 cm−1 [31]. The absorption peak of insulin had no change when the concentration of Fe3O4 NPs changed, suggesting that these Fe3O4 NPs purely adsorbed on the surface of insulin fibrils, and would not affect the fibrillation and any secondary structural changes of insulin fibrils.
Figure 5. (a) CD spectra of insulin fibrillation in the different contractions of Fe3O4 before incubation at pH 1.6 and 80 °C for 2.5 h (b) FTIR spectra of insulin fibrillation in the different contractions of Fe3O4 before incubation at pH 1.6 and 80 °C for 2.5 h.
Download figure:
Standard image High-resolution imageFurthermore, the TEM images indicated that a significant increase in Fe3O4 NPs bound to insulin fibrils were observed when insulin fibrils incubated with more amounts of Fe3O4 NPs, as depicted in figure 6. When 600 ppm Fe3O4 NPs were added in an aqueous solution, more of the Fe3O4 NPs bound to insulin fibrils (figure 6(b)). When the amount of Fe3O4 NPs was further increased to 1000 ppm, all insulin fibers were covered with Fe3O4 NPs (figure 6(c)).
Figure 6. TEM images of (a) 200 ppm, (b) 600 ppm, and (c) 1000 ppm of Fe3O4 bound to insulin fibrils incubation in a solution at pH 1.6 and 80 °C for 2.5 h, and then stored at room temperature for 3 weeks.
Download figure:
Standard image High-resolution image4. Conclusions
This study described the successful synthesis and characterization of magnetic Fe3O4 NPs and outlined the fact that Fe3O4 NPs had no effect on the conformational change of insulin fibrils, even with an increase in the concentration of Fe3O4 NPs, indicating the interaction between Fe3O4 NPs and insulin fiber was physical adsorption. Our results showed that Fe3O4 NPs did not induce the aggregation of insulin fibrils. The zeta potential of insulin fibrils and Fe3O4 were positive, suggesting the binding forces between Fe3O4 NPs and insulin fibrils were van der Waals forces (figure 7). Moreover, the removal of the insulin amyloid fibrils from its continuous phase by magnetization revealed that Fe3O4 NPs bound to insulin fibrils during the fibril reaction. Our data suggested that Fe3O4 NPS could be a potential clinical application for the accumulation of amyloid fibrils.
Figure 7. A scheme of Fe3O4 NPs adsorbed on insulin fibrils via van der Waals forces.
Download figure:
Standard image High-resolution image
Acknowledgments
The authors would like to thank the Ministry of Science and Technology of Taiwan (NSC 101-2113-M-110-013-MY3), (MOST 104-2628-M-110-001-MY3) and (MOST 103-2320-B-006-025-MY3), KMU-TP103G00, KMU-TP103G03, and the National Sun Yat-Sen University Biochip Research Group and Center for Neuroscience for financial support of this work.
Additional Information
The authors declare no competing financial interest.