Dihydrocaffeic Acid-Decorated Iron Oxide Nanomaterials Effectively Inhibit Human Calcitonin Aggregation.

To date, more than 30 human peptides or proteins have been found to form amyloid fibrils, most of which are associated with human diseases. However, currently, no cure for amyloidosis exists. Therefore, development of therapeutic strategies to inhibit amyloid formation is urgently required. Although the role of some amyloidogenic proteins has not been identified in certain diseases, their self-assembling behavior largely affects their bioactivity. Human calcitonin (hCT) is a hormone peptide containing 32 amino acids and is secreted by the parafollicular cells of the thyroid gland in the human body. It can regulate the concentration of calcium ions in the blood and block the activity of osteoclasts. Therefore, calcitonin has also been considered a therapeutic peptide. However, the aggregation of hCT hinders this process, and hCT has been replaced by salmon calcitonin in drug formulations. Recently, iron oxide nanomaterials have been developed as potential materials for various applications owing to their high biocompatibility, low toxicity, and ease of functionalization. In this study, nanoparticles (NPs) were prepared using a simple chemical coprecipitation method. We first demonstrated that dopamine-conjugated Fe3O4 inhibited hCT aggregation, similar to what we found when carbon dots were used as core materials in the previous study. Later, we continued to simplify the preparation process, that is, the mixing of dihydrocaffeic acid (DCA) and iron oxide NPs, to maintain their stability and inhibitory effect against hCT aggregation. Furthermore, DCA-decorated Fe3O4 can dissociate preformed hCT amyloid fibrils. This appears to be one of the most promising ways to stabilize hCT in solution and may be helpful for amyloidosis treatment.

Introduction

Amyloids are protein aggregates with a fibrillar morphology. X-ray diffraction studies revealed that these fibrils have a “cross-β” sheet structure and can be identified using particular dyes, such as Congo red or thioflavin-S.[1,2] Amyloid formation has been linked to the development of various diseases, including Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, and type II diabetes.[3,4] Human calcitonin (hCT), a 32-residue hormone peptide secreted from the parafollicular cells of the thyroid gland, is also an amyloidogenic protein. However, the aggregated form of hCT has not been universally associated with a particular disease. The physiological function of this hormone peptide is to regulate calcium and phosphate concentrations in the blood. It can block the activity of osteoclasts and decrease calcium resorption in the kidneys.[5] Moreover, hCT is secreted in response to high serum blood calcium levels. Pharmacological doses of calcitonin have been used to decrease bone resorption in osteoporosis, Paget’s bone disease, and hypercalcemia of malignancy.[6,7] However, hCT has been currently replaced by salmon calcitonin in drug products because of the aggregation propensity of hCT in aqueous solutions.[8,9] The development of novel excipients that would stabilize hCT would greatly enhance its use in therapeutic treatments.

Nanoparticles (NPs) are widely used in various biomedical applications. Some research groups have reported the use of NPs as a therapeutic approach to target amyloidosis. Recent data suggest that these nanomaterials can act as potential inhibitors of amyloid aggregation.[10−15] Moreover, it was observed that NPs had the ability to affect amyloid formation owing to their surface properties. Surface-coated NPs were first developed to increase their bioactivity and biocompatibility. Because the interaction of NPs with proteins influences protein structure and function, surface functionality seems to largely dominate their effect on protein aggregation. For example, amino acids are common coating reagents that improve the biological compatibility of NPs. It was found that three different amino acids (Gly, Lys, and Trp)-coated iron oxide NPs greatly affect the amyloid formation of lysozyme compared to uncoated NPs.[16] Among these, the most effective is the Trp-coated nanomaterial. It was assumed that more monomeric proteins would be adsorbed to reduce the number of aggregation-prone species in the solution. Dextran-based polymers have been modified and coated on magnetic NPs in an attempt to obtain efficient amyloid suppressors. For human insulin, diethylaminoethyl-dextran (DEAE-Dex)-coated NPs exhibited the best inhibition.[12] However, when the same nanomaterials were examined to prevent lysozyme amyloid formation, DEAE-Dex showed the lowest tendency to inhibit lysozyme fibril formation; instead, carboxymethyl-dextran was more effective in this case.[17] Although the size of the NPs is dependent on the type of coating agent and may alter the inhibitory effects of NPs, it is believed that the surface characteristics of NPs are crucial for protein–NP interactions. NP characteristics such as a large surface area and appropriate surface modification would actively block fibril formation due to protein–NP interaction.

We have previously demonstrated that dopamine-conjugated carbon dots effectively inhibit hCT fibrillization.[18] Moreover, this nanomaterial with a size of less than 10 nm can also dissociate preformed hCT amyloids. The synthetic approach to creating carbon dots is simple and eco-friendly. However, calcination of the carbon source requires high heat. A standard coprecipitation method for preparing iron oxide NPs is an easier approach.[19−21] Here, we examined the influence of dopamine-conjugated iron oxide NPs (DO–Fe3O4) on hCT aggregation. It was found that DO–Fe3O4 was able to inhibit hCT amyloid formation. However, it was not as efficient as using carbon dots as a core material, and the stability of the catechol-functionalized nanomaterials also affected its antiamyloid activity. Therefore, we prepared dihydrocaffeic acid (DCA) coated iron oxide NPs (DCA@Fe3O4) for the same purpose. Our aim was to develop an easy method of preparation of nanomaterials for blocking protein aggregation, and we believe that this would facilitate the use of hCT in drug formulations.

Materials and Methods

Peptide Synthesis and Purification

hCT and IAPP in this study were both synthesized through a Liberty Lite microwave-assisted solid-phase peptide synthesizer with fluorenylmethyloxycarbonyl chemistry. A detailed description of peptide synthesis and purification has been described elsewhere.[18,22] Briefly, a low-loading Rink Amide ProTide Resin (CEM Corporation, 0.16 mmol/g) was used to work as solid-phase support and provided amidated C-terminus. 10% piperazine (w/v) in a solution of ethanol and N-methylpyrrolidone (10:90) was prepared as deprotection reagents and diisopropylcarbodiimide prepared in dimethylformamide worked as activation reagents in each cycle. Once the synthesis is complete, peptides were cleaved from the resin using a cleavage cocktail including trifluoroacetic acid, water, triisopropylsilane, and 3,6-dioxa-1,8-octanedithiol (92.5:2.5:2.5:2.5). Both peptides were further oxidized using I2 dissolved in methanol to form disulfide bonds. After purification via reverse-phase high-performance liquid chromatography, peptides were collected and subjected to mass spectrometry to confirm the molecular weight.

Protein Sample Preparation

After purification, peptides were dispensed into microtubes and saved in powder form at −20 °C. Before experiments, protein powder was first treated with pure hexafluoro-2-propanol (1 μg/μL) for 5–6 h to dissociate any potential peptide aggregates and then lyophilized overnight. The resulting hCT peptide powder was first prepared in 300 μL, 50 mM phosphate buffer (pH 7.4), and IAPP peptide powder was prepared in 300 μL, 10 mM Tris buffer (pH 7.4), respectively. Peptide solutions were centrifuged at 15,000 rpm for 10 min to remove any insoluble aggregates. 10 μL of each peptide stock solution was used to determine protein concentration using a bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, USA) according to user manuals, and later, the rest was diluted to the desired concentration by an appropriate buffer.

Preparation and Characterization of Iron Oxide Nanomaterials

In this study, we tested three kinds of iron oxide NPs. Dopamine-conjugated Fe3O4 (noted as Do–Fe3O4) was prepared via the coupling reaction of dopamine with COOH–Fe3O4. COOH–Fe3O4 was prepared via the coupling reaction of sodium alginate with NH2–Fe3O4. The preparation of COOH–Fe3O4 and NH2–Fe3O4 have been previously described.[23,24] In brief, dopamine (4 mg) was covalently conjugated onto COOH–Fe3O4 (2 mg) by using cross-linking reagent EDC·HCl (3.9 mg) and NHS (2.4 mg) in 5 mL deoxygenated MES buffer at pH 5.0. The solution was then stirred, with nitrogen being bubbled through, for 24 h. Later, the supernatant was removed by centrifugation at 15,000 rpm for 10 min, and the NPs were further washed three times with deionized water. On the other hand, DCA@Fe3O4 was afforded via incubation of DCA (53 mg) with Fe3O4 NPs (18 mg) produced by the methods of coprecipitation in 1.25 mL phosphate-buffered saline (PBS). First, 1 M ferric chloride hexahydrate (FeCl3·6H2O) and 2 M ferrous chloride tetrahydrate (FeCl2·4H2O) were prepared by dissolving iron salts separately in 2 M HCl solutions. 4 mL of 1 M FeCl3 solution was mixed with 1 ml of 2 M FeCl2 solution without the addition of glycine in a flask. The solution was vigorously stirred, and then, 5 M NaOH solution was added very slowly until the solution turned black. The precipitates were collected by a permanent magnet and washed twice with deionized water. DCA and bare Fe3O4 were mixed for 2 h and followed by 30 mins of sonication. Again, the supernatant was removed by centrifugation at 15,000 rpm for 10 min and the NPs were further washed three times with deionized water.

The hydrodynamic size distribution and the zeta potential of Do–Fe3O4 and DCA@Fe3O4 were determined using a dynamic light scattering (DLS)-based particle size analyzer (Otsuka Electronics, Osaka, Japan) in deionized water at 25 °C. The Fourier transform infrared (FTIR) spectra were recorded to identify the functional groups of DO–Fe3O4 and DCA@Fe3O4 (Spectrum 500, PerkinElmer, Waltham, MA, USA). The composition of the NPs was confirmed by energy-dispersive X-ray spectroscopy (EDX) incorporated into a scanning electron microscopy system (Hitachi S-4800, Japan).

Thioflavin-T Kinetic Assay

The thioflavin-T (ThT) assay was used to measure the kinetics of hCT amyloid formation alone and in the presence of Do–Fe3O4 and DCA@Fe3O4. The peptide solution, in general, was prepared in 50 mM PB buffer with 16 μM ThT and three concentrations of NPs (50, 100, and 200 μg/mL) were applied. Assays were performed at 25 °C in a sealed 384-well nonbinding surface microplate (40 μL loading volume per well) with agitation every 1 h. For fibril dissociation experiments, fibril solution was afforded by incubating 64 μM hCT monomers in microtubes with continuous shaking for more than 3 days. Later, hCT amyloid fibrils mixing with ThT and NPs or buffer (control groups) were dispensed into each 384-well microplate. The concentration would be down to 57 μM due to the addition of buffer or NPs. Measurements were made using a multimode microplate reader (SpectraMax M2, Molecular Devices, Sunnyvale, CA, USA) with excitation at 430 nm and emission at 485 nm. In order to ensure experimental consistency, each condition was tested at least using three batches of sample preparation.

Transmission Electronic Microscopy

The morphology of NPs or the end-products of ThT assays were examined using a Hitachi H-7100 transmission electron microscope with an accelerating voltage of 120 kV. The protein sample was further stained with 2% uranyl acetate to enhance contrast. This experiment was performed in an instrumentation center at National Taiwan University.

Fluorescence Microscopy

10 μL of the end-products from ThT assays were taken out from a 384-well microplate and immobilized on a thin layer of 1% agarose in PBS onto the clean microscopic slide individually. Phase contrast and fluorescence microscopy images were obtained using a Zeiss Axio A1 Microscope with an EC Plan-Neofluar 100×/1.3 oil objective lens.

4,4′-Dianilino-1,1′-binaphthyl-5,5′-disulfonic Acid (Bis-ANS) Binding Assay

A fresh stock solution of 0.5 mM Bis-ANS was first dissolved in 50 mM PB buffer. To further confirm the inhibitory effect of iron oxide NPs in hCT fibril formation, we prepared 64 μM hCT monomers with 200 μg/mL NPs in microtubes and also have 64 μM hCT monomers alone as a control study. Samples were incubated at 25 °C for 24 h with mild agitation at 300 rpm. Later, Bis-ANS was added to individual microtubes and subjected to a microplate reader to perform fluorescence measurement. The emission spectra were collected from 450 to 550 nm upon excitation at 355 nm.

Circular Dichroism

The sample preparation for the circular dichroism (CD) experiment was similar to that for Bis-ANS binding assays. CD measurements were conducted after 0, 6, 18, 26, and 45 h incubation using a J-715 circular dichroism spectrometer (JASCO, USA). Spectra were recorded from 200 to 250 nm using a 1 mm path length quartz cell at 25 °C (1 nm intervals; scan speed 50 nm/min). The data were averaged from 10 scans and subtracted from the background spectrum.

Results and Discussion

Catechol Functional Groups Coated with Iron Oxide NPs Were Synthesized and Characterized

In previous studies, polyphenolic compounds, including epigallocatechin gallate, quercetin, myricetin, and baicalein, with catechol groups, have been shown to inhibit hCT aggregation.[25,26] Moreover, we conjugated dopamine molecules on the surface of the carbon dots via the formation of covalent bonds with the amino groups of dopamine. The exposure of multiple catechol groups on carbon dots allows this water-soluble nanomaterial to effectively inhibit hCT fibrillization and disassemble preformed amyloid fibrils.[18] Similar approaches were used in this study to synthesize DO–Fe3O4. Irrespective of the dopamine-conjugated carbon dots or DO–Fe3O4, the stability of NPs is an important issue that needs to be addressed to achieve better efficacy. With recent advances in surface chemistry for the fabrication of functional groups on iron oxide NPs,[27] it is possible to form extremely stable complexes of catechol ligands and carboxylic acid-based molecules with iron oxide NPs via an adsorption mechanism. Thus, we used a derivative of dopamine, DCA, to form a surface coating on iron oxide NPs to enhance their stability and maintain their antiamyloid properties. The procedures for the preparation of DO–Fe3O4 and DCA@Fe3O4 are illustrated in Scheme . The properties of DO–Fe3O4 and DCA@Fe3O4 were characterized using several methods. DLS measurements were performed to evaluate the hydrodynamic diameter of the nanomaterials. As presented in Figure and Table , the hydrodynamic diameters of the five examined nanomaterials were determined between 50 and 100 nm. The size and morphology of the nanomaterials determined by transmission electronic microscopy (TEM) were ∼5 nm and similar to each other. In addition, the zeta potential of the nanomaterials revealed apparent differences owing to their surface characteristics. NH2–Fe3O4 was positively charged in water. The Z values of NH2–Fe3O4 and DO–Fe3O4 were −30.0 ± 0.9 and −21.0 ± 1.1 mV, respectively. The Z value of DO–Fe3O4 was higher than that of COOH–Fe3O4, and the oxygen and nitrogen element components determined using EDX analysis (Table ) increased for DO–Fe3O4 compared to those for COOH–Fe3O4, suggesting that dopamine conjugation was successful, although the FTIR spectra of DO–Fe3O4 were similar to the spectra recorded for NH2–Fe3O4 and COOH–Fe3O4 (Figure S1). The spectra of NH2–Fe3O4 were in line with the previous data.[23] On the contrary, the Z values of bare Fe3O4 and DCA–Fe3O4 were 35.7 ± 0.5 and −31.6 ± 0.5 mV, respectively. The negatively charged DCA–Fe3O4 strongly reversed the electronic properties of bare Fe3O4, indicating that DCA was adsorbed on the surface of Fe3O4. EDX analysis also indicated that the percentage of iron atoms decreased after compound absorption. A much stronger C=O stretching vibration band at υ̃ = 1642 cm–1 was observed in the FTIR spectra of DCA@Fe3O4. We also performed a simple test to understand the thermal stability of DCA@Fe3O4 because sometimes the COOH coordination bond is labile and could be easily broken by an increase in temperature. DCA@Fe3O4 was prepared at a concentration of 200 μg/mL in water and was heated at 37 °C for 1 h. The zeta-potential results for DCA@Fe3O4 remained similar before and after heating (Figure S1c).

Scheme 1

Synthetic Procedure for the Preparation of DO–Fe3O4 and DCA@Fe3O4

Figure 1

(a) Size distribution of NPs determined by DLS measurements. (b) TEM images of NPs. Scale bars represent 50 nm.

Table 1

Characteristics of DO–Fe3O4, DCA@Fe3O4, and Their Precursors

	NH2–Fe3O4	COOH–Fe3O4	DO–Fe3O4	Fe3O4	DCA@Fe3O4
Zeta (mV)	44.5 ± 1.2	–30.0 ± 0.5	–21.0 ± 1.1	35.7 ± 0.5	–31.63 ± 0.5
DLS (nm)	58.2 ± 7.3	64.6 ± 17.8	83.4 ± 32.0	76.8 ± 21.6	57.9 ± 6.5
TEM (nm)	5.7 ± 1.5	5.7 ± 1.9	6.3 ± 1.6	4.9 ± 0.7	6.0 ± 1.5

Table 2

Characteristics of DO–Fe3O4, DCA@Fe3O4, and Their Precursors

	NH2–Fe3O4		COOH–Fe3O4		DO–Fe3O4		Fe3O4		DCA@Fe3O4
element	mass (%)	atom (%)	mass (%)	atom (%)	mass (%)	atom (%)	mass (%)	atom (%)	mass (%)	atom (%)
N	3.40	5.91	4.29	6.84	4.81	7.46	N/Aa	N/A	N/A	N/A
O	47.79	72.79	55.21	76.98	57.40	77.86	46.61	75.29	73.78	90.76
Fe	48.81	21.3	40.50	16.18	37.79	14.68	53.39	24.71	26.22	9.24

N/A: not applicable.

DCA@Fe3O4 Is a More Efficient Inhibitor of hCT Aggregation Than DO–Fe3O4

We investigated the inhibitory effect of DO–Fe3O4 and DCA@Fe3O4 on hCT fibrillation by monitoring the kinetics of hCT amyloid formation using ThT assays and further validated the ThT results using TEM. Besides, we also tested small molecules, dopamine and DCA as control studies. The fluorescent dye ThT has become the most widely used classic probe for selectively identifying β-rich structured amyloid fibrils since its application was described in 1959.[28] Although many fluorescent probes have been developed to replace ThT, only a few have been commercialized. In general, ThT fluorescence showed a typical sigmoidal curve during the protein incubation time. Initially, during a time period, called the lag phase, ThT was undetectable because these small aggregates were still not β-sheet dominant. Later, a significant ThT enhancement was detected during the fibril-forming growth phase until a plateau was formed in which monomers, oligomers, and fibrils reached a new equilibrium.[29] From our observations, the lag time of hCT fibrillization was approximately 18 h (black curve in Figure a–c). Dopamine and DCA are indeed effective in preventing hCT aggregation (Figure S2). However, these two small molecules are easily oxidized in an aqueous solution and lose their efficacy within 1 day. The addition of bare Fe3O4 to the hCT aggregates had trivial effects. The hCT amyloid fibrils with typical morphology were observed after 3 days of incubation. DO–Fe3O4 at a higher concentration (200 μg/mL) seemed to suppress the formation of amyloid fibrils, leading to a much lower ThT intensity, suggesting that catechol functionalization was indeed a key component in preventing hCT fibril formation. In contrast, we found that DCA@Fe3O4 behaved more like dopamine-conjugated carbon dots and more effectively inhibited hCT fibril formation. Although the lag time of hCT in the presence of DCA@Fe3O4 did not extend, the ThT signals began to increase after 20 h and stopped after 5–6 h of incubation. The final ThT intensity also correlated with the amount of DCA@Fe3O4 initially applied. The inhibitory effect of DCA@Fe3O4 on hCT fibrillization was concentration-dependent. TEM images also verified that the hCT samples with 200 μg/mL DO–Fe3O4 and DCA@Fe3O4 did not form a large quantity of amyloid fibrils. Moreover, we used fluorescence microscopy to image the amyloid fibrils of hCT (Figure S3). Only hCT samples showed fluorescence fibrillar structures. Other hCT samples coincubated with nanomaterial exhibited loose amorphous aggregates. Thus, nonamyloid aggregates may be coprecipitates of NPs and proteins. The NPs did not allow ThT binding; thus, there was no fluorescence signal.

Figure 2

Effects of iron oxide NPs on the amyloid formation of hCT in vitro. ThT kinetics assays monitored for hCT fibrillization in the presence of (a) bare Fe3O4, (b) DO–Fe3O4, and (c) DCA@Fe3O4. hCT was prepared at a concentration of 64 μM in 50 mM phosphate buffer at pH 7.4. Three different concentrations (50, 100, and 200 μg/mL from light to dark color) of NPs were tested. Experiments were conducted in triplicate. TEM images were recorded for hCT with 200 μg/mL (d) bare Fe3O4, (e) DO–Fe3O4, and (f) DCA@Fe3O4 after 72 h incubation.

Moreover, we performed another aggregation assay by preparing hCT samples and iron oxide NPs in microtubes. The samples were incubated at 25 °C for 24 h with mild agitation at 300 rpm. Subsequently, we analyzed the final states of the hCT samples. First, we filtered the samples after 24 h incubation to determine the concentration of the remaining protein monomers in the final solution using the protein BCA assay. The absorbance measured at 562 nm indicated that hCT with DCA@Fe3O4 coincubation showed the highest value, suggesting that more protein monomers were present in the solution (Figure a). On the contrary, the same half samples were probed by adding Bis-ANS dye. Bis-ANS is a fluorescent probe mainly used for detecting nonpolar cavities in proteins. In a previous study, Bis-ANS and ANS were compared to ThT for their ability to monitor amyloid-beta (Aβ) and islet amyloid polypeptide (IAPP) amyloid formation.[30] Bis-ANS has been demonstrated to be capable of determining the kinetics of Aβ and IAPP aggregation. According to the test results, Bis-ANS was more sensitive to fibril detection than ThT and had a higher binding affinity for Aβ fibrils than ThT. Although we could not confirm whether Bis-ANS had a higher preference for binding to hCT amyloid fibrils than ThT, we did confirm that Bis-ANS can probe hCT amyloid fibrils by exhibiting fluorescence enhancement at 490 nm compared to the probe alone. However, Bis-ANS did not show similar fluorescence signals when applied to hCT samples with iron oxide NPs (Figure b). Both the assays supported the data obtained from the ThT experiments. The two types of NPs synthesized in this study could effectively inhibit hCT fibrillization.

Figure 3

Final states of hCT amyloid formation with and without iron-oxide NPs were checked by (a) protein BCA assay and (b) Bis-ANS fluorescence assay. The black bar or curve represents hCT samples only. The green bar or curve represents hCT samples with 200 μg/mL DO–Fe3O4. The blue bar or curve represents hCT samples with 200 μg/mL DCA@Fe3O4.

DO–Fe3O4 and DCA@Fe3O4 Inhibited the Nucleation and Elongation State of hCT Fibrillization

We examined the inhibitory effect of DO–Fe3O4 and DCA@Fe3O4 during nucleation and elongation states of hCT fibrillization. Based on the previous data, we found that DO–Fe3O4 and DCA@Fe3O4 cannot modulate the lag time of hCT amyloid formation; this may indicate that the interactions of protein monomers and NPs might be weak. In order to understand the details of the inhibitory mechanism of DO–Fe3O4 and DCA@Fe3O4, we added DO–Fe3O4 and DCA@Fe3O4s before and during the fibril growth state of hCT fibrillization. Based on the previous ThT kinetic results shown in Figure , DO–Fe3O4 and DCA@Fe3O4 were added at 20 and 30 h after the hCT fibrillization had started. Buffer was added instead of DO–Fe3O4 and DCA@Fe3O4 in the control study. Post addition of buffer at 20 h, the lag time of hCT aggregation alone increased to 30 h (Figure a). However, the addition of bare Fe3O4 at 50 or 200 μg/mL had no effect on lag time extension or the final intensity of ThT. However, the addition of DO–Fe3O4 and DCA@Fe3O4 at 20 h led to an increase in the lag time of hCT aggregation by 45 h or even longer. The ThT final intensity decreased significantly, although only 50 μg/mL of DO–Fe3O4 was added. Both DO–Fe3O4 and DCA@Fe3O4 still exhibited concentration-dependent inhibitory effects, and DCA@Fe3O4 was still more effective than DO–Fe3O4. By contrast, the final ThT signals were found to be even lower when iron oxide NPs were added during the hCT fibril elongation phase when compared to each condition recorded in Figure a (Figure b). In some cases (addition of buffer, bare Fe3O4, and 50 μg/mL DO-Fe3O4), the ThT intensity decreased after the addition of buffer or nanomaterials but gradually increased and reached a plateau. However, in other cases (addition of 200 μg/mL DO–Fe3O4 and DCA@Fe3O4), the ThT intensity continued to decrease within 3 h and did not increase. These data, combined with recorded TEM images (Figure c–f), suggest that DCA@Fe3O4 and DO–Fe3O4 at high concentrations can not only suppress fibril growth but also disassemble protofibrils. Moreover, the addition of bare Fe3O4 seemed to have a minor effect on disrupting fibril growth. This is different from our observation when the nanomaterials were introduced at the initial or nucleation stage of hCT aggregation.

Figure 4

Effects of iron oxide NPs during nucleation and the elongation state of hCT fibrillization. The hCT aggregation kinetics monitored by ThT assay. Iron-oxide NPs were added at (a) 20 and (b) 40 h (arrow indicated). Black: buffer; dark grey: 200 μg/mL bare Fe3O4; light grey: 50 μg/mL bare Fe3O4; dark green: 200 μg/mL DO–Fe3O4; light green: 50 μg/mL DO–Fe3O4; dark blue: 200 μg/mL DCA@Fe3O4; and light blue: 50 μg/mL DCA@Fe3O4. TEM images for ThT end products with the addition of (c) buffer and 200 μg/mL (d) bare Fe3O4, (e) DO–Fe3O4, and (f) DCA@Fe3O4 at 40 h. Scale bars represent 200 nm.

Iron Oxide NPs Effectively Reversed or Destroyed hCT and IAPP Amyloid Fibrils

Although many aspects of NPs with different surface properties have been applied to perturb amyloidogenic protein aggregation, few examples have demonstrated positive effects on the disassembly of preformed amyloid fibrils. Tomašovičová et al. used atomic force microscopy to study the interaction between lysozyme amyloid fibrils and assorted diameters.[31] The NPs immediately agglomerated onto the fibrillary surface. Approximately, 10 nm NPs completely disintegrated the lysozyme amyloid fibrils into fragmented debris. Based on the observation shown in Figure b, whether the nanomaterials disrupted the mature fibrils need to be studied. Therefore, we first prepared hCT amyloid fibrils from protein monomers and incubated the hCT samples in microplates. The ThT fluorescent dye was applied again to confirm fibril formation. At 90 h, the buffer and nanomaterials were added, and ThT signals were recorded every 10 min (Figure a). After buffer addition, the ThT intensity gradually decreased upon the addition of NPs, including bare Fe3O4. The extent of the decrement was in the order of DCA@Fe3O4 > DO–Fe3O4 > bare Fe3O4. TEM was performed to record the status of the hCT samples collected at 100 h. A large amount of fibrillary hCT amyloid formation after buffer addition was observed (Figure b). However, bare Fe3O4 changed the fibril morphology of hCT, and several amorphous aggregates were observed in the TEM images (Figure c). Compared to bare Fe3O4, DO–Fe3O4 and DCA@Fe3O4 were more effective in dissociating mature fibrils. Although similar tests were performed in a previous study,[18] and the effectiveness of dopamine-conjugated carbon dots and Fe3O4 were not directly compared; here, DO–Fe3O4 appeared to perform better. Several minute fragments were only found in the hCT fibril samples upon the addition of DO–Fe3O4 (Figure d). DCA@Fe3O4 was better at disaggregating hCT amyloid. A few aggregates are observed in Figure f. The intensity of ThT decreased significantly in the first couple of minutes suggesting that the disaggregation of hCT amyloid by DCA@Fe3O4 was very fast. Although a time delay would be produced due to the addition of all the experimental samples into the microplate before the first reading of the ThT signal, such a timescale of dissociation behavior is possible. Noi et al. successfully monitored the disassembly of the Aβ fibrils by anthocyanins using total internal reflection fluorescence microscopy (TIRFM).[32] Decomposition was complete within ∼2 h and some fibrils were decomposed in the first 20 min, as revealed by time-course TIRFM measurements. Thus, the dissociating ability of DCA@Fe3O4 on other amyloidogenic peptides was also studied. Here, we prepared IAPP fibrils, incubated them with 50 or 200 μg/mL DCA@Fe3O4 for 10 h, and monitored the change in ThT intensity during this period. Based on the rate at which the ThT signal decreased and the TEM images, it was understood that DCA@Fe3O4 can dissociate IAPP fibrils; however, the time taken was longer than that taken to dissociate hCT (Figure S4a).

Figure 5

Iron oxide NPs disrupted preformed hCT fibrils to different extents. (a) hCT fibrils dissociation kinetics monitored by ThT assay. Black: buffer; dark grey: 200 μg/mL bare Fe3O4; light grey: 50 μg/mL bare Fe3O4; dark green: 200 μg/mL DO–Fe3O4; light green: 50 μg/mL DO–Fe3O4; dark blue: 200 μg/mL DCA@Fe3O4; and light blue: 50 μg/mL DCA@Fe3O4. TEM images for hCT fibrils with the addition of (b) buffer and 200 μg/mL (c) bare Fe3O4, (d) DO–Fe3O4, and (e) DCA@Fe3O4 after 10 h of incubation. Scale bars represent 200 nm.

DO–Fe3O4 and DCA@Fe3O4 Exhibit Stronger Interaction with hCT

The time-course of hCT with iron oxide NPs and fibril formation was monitored using CD at 0, 6, 18, 26, and 45 h. This allowed us to understand the conformational change of hCT modulated by the inhibitory property of DO–Fe3O4 and DCA@Fe3O4. The samples in microtubes were incubated while being subjected to mild agitation. At the desired time points, the samples were transferred to a quartz cuvette and subjected to CD measurements. Both hCT alone and hCT with DO–Fe3O4 and DCA@Fe3O4 showed a predominant random coil conformation in the first 6 h. With respect to the sample containing hCT alone, the gradual loss of the CD signals suggested the formation of calcitonin amyloids, which was confirmed using TEM (Figure S5); however, this phenomenon is not typical for other amyloidogenic proteins. Protein precipitates may be the reason for the loss of CD signals. However, we found similar results for hCT incubated with DO–Fe3O4 and DCA@Fe3O4, in contrast to our observations from the ThT assays and TEM, which indicated that DO–Fe3O4 and DCA@Fe3O4 inhibited hCT amyloid formation (Figure a–c). Therefore, we measured the surface zeta potential of the NPs, including bare Fe3O4 in phosphate buffer, to understand the physical stability of the nanosuspensions in the presence of hCT. Bare Fe3O4 showed the same measured zeta-potential results both with and without incubation with hCT. DO–Fe3O4 and DCA@Fe3O4 showed increased zeta-potential values in the presence of hCT compared to that of NPs alone in phosphate buffer, suggesting protein binding to the surface of functionalized iron oxide NPs (Figure d) that consequently interrupt the interactions among hCT.

Figure 6

DO–Fe3O4 and DCA@Fe3O4 may have stronger binding affinity to hCT monomers than bare Fe3O4. Time course CD spectra of (a) hCT alone, (b) hCT with DO–Fe3O4, and (c) hCT with DCA@Fe3O4. hCT was prepared at 64 μM in 50 mM PB buffer with 200 μg/mL NPs. Measurements were conducted after 0, 6, 18, 26, and 45 h incubation. (d) Zeta potential measurements for iron-oxide NPs in buffer and in the presence of hCT. Solid bars represent the results for NPs in buffer. Patterned bars represent the results for NPs in the presence of hCT.

Conclusions

The data shown here demonstrate that DO–Fe3O4 and DCA@Fe3O4 effectively inhibit hCT amyloid formation, especially DCA@Fe3O4. Moreover, the preparation of DCA@Fe3O4 is simple and straightforward. The two iron oxide NPs were found to interrupt amyloid formation, even when hCT was associated and became oligomers or protofibrils. Although bare Fe3O4 did not inhibit hCT aggregation very well owing to its low binding affinity to protein monomers, it can moderately dissociate preformed hCT amyloid fibrils. Stability is critical for all drug products. The use of hCT as an active pharmaceutical ingredient is difficult because of its strong aggregation property. The development of hCT variants with low aggregation propensity is an alternative approach. However, the bioactivity of hCT variants should be further examined to meet the need for therapeutic treatment. The selection of excipients in the final formulation to achieve protein stabilization is common but challenging. We examined iron oxide NPs with appropriate decoration to prevent hCT aggregation. This information would be beneficial for drug formulation.