κ-Casein-Based Hierarchical Suprastructures and Their Use for

Article pubs.acs.org/Biomac

κ‑Casein-Based Hierarchical Suprastructures and Their Use for Selective Temporal and Spatial Control over Neuronal Differentiation Jiyeong Chun, Ghibom Bhak, Sang-Gil Lee, Ji-Hye Lee, Daekyun Lee, Kookheon Char, and Seung R. Paik* School of Chemical and Biological Engineering, Institute of Chemical Processes, College of Engineering, Seoul National University, 599 Gwanak-Ro, Gwanak-Ku, Seoul, Korea, 151-744 S Supporting Information *

ABSTRACT: Functions are diversified by producing hierarchical structures from a single raw material. Biologically compatible milk protein of κ-casein has been employed to fabricate higher-order suprastructures. In the presence of dithiothreitol and heat treatment, κ-casein transforms into amyloid fibrils with distinctive morphology attributable to mechanism-based fibrillar polymorphism. As the fibrils elongate to yield high aspect ratio during high-temperature incubation, the resulting fibrils laterally associate into the liquid crystalline state by forming a two-dimensional fibrillar array. Following a desalting process, the fibrillar arrays turn into a three-dimensional matrix of hydrogel that could be selectively disintegrated by subsequent salt treatment. The hydrogel was demonstrated to be a matrix capable of exhibiting controlled release of bioactive substances like retinoic acid, which led to temporal and spatial control over the differentiation of neuronal cells. Therefore, the hierarchical suprastructure formation derived from the single protein of κ-casein producing one-dimensional protein nanofibrils, a two-dimensional liquid crystalline state and a three-dimensional hydrogel could be widely appreciated in various areas of nanobiotechnology including drug delivery and tissue engineering.

■

INTRODUCTION Protein-based suprastructures are considered to be versatile nanoscale materials particularly useful at the biological interface as their chemical natures could be easily modified. Their intrinsic biocompatibility allows them to be the junctional materials possibly employed between living cells and inanimate objects.1−4 Amyloid fibrils are ordered insoluble protein nanofibrils with a width of 10−20 nm generated via selfassembly of various soluble amyloidogenic proteins.5 These fibrils are stabilized by the common secondary structure of cross β-sheet conformation,6 which might be responsible for their mechanical strength comparable to spider silk.7 Amyloid fibrils have been a pathological target of intensive research as they are involved in numerous neurodegenerative disorders including Alzheimer’s, Parkinson’s, and Prion diseases,8−11 although their causal role for the cell death remains to be unequivocally elucidated.9 Natural amyloids have been demonstrated to exhibit several biological functions such as biofilm12,13 and hyphae formation of microorganisms,14,15 egg envelope formation of insects and fish,16,17 and melanin biosynthesis for mammals.18 Materialistic value of the amyloid fibrils can be augmented as amyloidogenic proteins/peptides are engineered to produce © 2012 American Chemical Society

hierarchical suprastructures such as one-dimensional anisotropic nanofibrils, two-dimensional fibrillar arrays, and threedimensional fibrillar matrices. Although amyloidogenesis has been suggested to be a generic feature for proteins experiencing a partially misfolded state,5,11,19 the resulting amyloid fibrils apparently exhibit structural polymorphism depending on their constituting monomeric units and the assembly processes, which gives rise to distinctive fibrillar morphologies with respect to length, thickness, surface structure (smooth vs rugged), and overall shape (straight vs curly).20−24 In a previous study, thermally induced conformational change of βlactoglobulin led to different aggregation pathways, which could result in various end products.25 These polymorphic fibrils, therefore, increase their application potential, which might be further improved by additional physical/chemical modifications of the fibrils. For example, amyloid protein nanofibrils are suggested to be useful for fabricating functional nanomaterials such as conductive nanowires,26 metallic nanoparticle arrays,27 enzyme-immobilization matrix,28 protein liquid crystals,29 and Received: May 3, 2012 Revised: July 20, 2012 Published: July 31, 2012 2731

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738

Biomacromolecules

Article

water containing either 10 μM all-trans-retinoic acid (Sigma, R2625) or 50 μM FITC (Sigma, FD70). Instrumental Characterization of κ-Casein Fibrils and Hydrogel. For transmission electron microscope (TEM) analysis, the κ-casein fibrils were adsorbed onto a carbon-coated copper grid (Ted Pella Inc., U.S.A.) by placing 5 μL of the fibrillation mixture onto the grid and drying it by removing liquid with a filter paper. Following negative staining with 2% uranyl acetate, the fibrils were examined with TEM of JEM 1010 (JEOL, Japan). Field-emission scanning electron microscope (FE-SEM) images were obtained using SUPRA 55VP (Carl Zeiss, Germany) for the freeze-dried κ-casein hydrogel coated with platinum in 5 nm thickness using Sputter Coater SCD 005 (BALTEC). The hydrodynamic radius of κ-casein monomer/oligomer was determined by dynamic light scattering DLS-8000 (Otsuka Electronics Co. Ltd., Japan) at 20 °C. The vertically polarized 75 mW Ar laser (λo = 488 nm) was exposed to the sample and the resulting scattered light intensity was analyzed as a function of hydrodynamic radii of the particulates. To define the liquid crystal state of the κ-casein solution, the vials containing the amyloid fibrils of κ-casein (15 mg/mL) were located between two crossed polarizers in the dark. The light transmission images were taken by digital camera DSC-T33 (Sony, Japan). To evaluate disintegration of FITC-entrapped κ-casein hydrogel and its subsequent release of FITC, the gel was immersed in 10 mM sodium phosphate at pH 8.0 for 10 h in the diffusion-based gradient generator with a well-diameter of 5 mm, a side arm length of 600 μm, and a depth of 10 μm, which was prepared through rapid prototyping and soft lithography using poly(dimethylsiloxane) (PDMS). Fluorescence intensity of the released FITC was monitored with an inverted microscope of AE30 modified to detect the light emission (Motic, Hong Kong). Cell Study. The SH-SY5Y neuroblastoma cells were grown in Dulbecco’s modified Eagle’s medium (Hyclone) supplemented with 10% fetal bovine serum (GIBCO, 12662002) and penicillin/ streptomycin (GIBCO, 15140163) in a humidified incubator with 95% air and 5% CO2 at 37 °C. Cells were seeded with 1 × 106 cells/ plate onto a 9 cm culture dish and cultured for 20 h within the complete medium. Retinoic acid dissolved in DMSO at 10 μM was added to the cells where the final concentration of DMSO was maintained at 0.05%, a concentration that had no effect on cell growth and differentiation. The κ-casein hydrogel containing 10 μM all-transretinoic acid (Sigma, R2625) was placed at the designated corner of a 9 cm culture dish after 24 h of ultraviolet light treatment. The cells were examined with an inverted microscope IX71 (Olympus, Japan) to evaluate neurite extension in the dish. Mean percentage of the neurite outgrowth in comparison with control was obtained from the neuronal images. The average neuritic length was obtained by analyzing three different sets of neurons (∼100 cells/set) revealed in the microscopic images collected following 20 h of incubation. Statistical significance of the neurite outgrowth was assessed with Mann−Whitney rank sum test (*p < 0.05, **p < 0.002).

protein-based hydrogels.30−33 In fact, the nematic liquid crystal state as a product of the two-dimensional (2-D) fibrillar array has been derived from the amyloid fibrils of lysozyme as the protein was partially misfolded at low pH under heat treatment.34 Self-assembled peptides such as Aβ16−20 and diphenylalanine also led to a nematic liquid crystalline phase as they were treated with organic solvent or pH change.35,36 These 2-D arrays could provide a biologically controllable platform for the assembly of organic and inorganic materials.37 A threedimensional (3-D) matrix of amyloid fibrils has been generated in the predominant form of hydrogel. Recently, curly amyloid fibrils derived from α-synuclein, a pathological component of Parkinson’s disease, were demonstrated to produce a welldefined nanospaced transparent hydrogel in which chemicals, enzymes, and even cells could be entrapped and selectively released.24 Amyloidogenic peptides such as Lys β-21 and βamyloid diphenylalanine also produced hydrogels consisting of a fibrillar meshwork.36,38 Amyloid-based suprastructures, therefore, possess the potential to be applied in various areas such as biosensor development, drug delivery, and tissue engineering.39−42 κ-Casein is a stabilizing component of colloidal particles in milk solution, also known as casein micelles, comprising other additional proteins of αS1-casein, αS2-casein, and β-casein.43 κCasein itself, however, forms amyloid fibrils in vitro although the protein never transforms into the fibrils in the presence of the other caseins.43 Nontoxic, biocompatible, and chemically modifiable nature of κ-casein designates it an attractive raw material suitable for the functional suprastructure formation. In this report, we have demonstrated controlled assembly of κcasein to fabricate the hierarchical suprastructures of 1-D, 2-D, and 3-D protein-based biomaterials. In addition, selective disintegration of the final 3-D product of hydrogel allowed us to demonstrate temporal and spatial control over the differentiation of neuronal cells by providing a biocompatible matrix for controlled release of retinoic acid.

■

MATERIALS AND METHODS

Fibrillation of κ-Casein. Bovine milk κ-casein (Sigma, C0406) was dissolved in 10 mM sodium phosphate (pH 8.0) to 3 mg/mL and passed through 0.45 μm filters to remove undissolved precipitates. The concentration of κ-casein was adjusted to 1.5 mg/mL with the phosphate buffer in the presence or absence of 20 mM dithiothreitol (DTT) depending on fibrillation conditions. The protein mixtures were incubated at 37 °C and the fibrillation process was monitored by measuring dye-binding fluorescence using the amyloid fibril specific thioflavin-T. The κ-casein in 10 mM phosphate (pH 8.0) containing DTT was also incubated at various temperatures (20, 40, 60, 80 °C) under a quiescent condition. Aliquots from the fibrillation mixture containing 8 μM κ-casein were combined with 2.5 μM thioflavin-T in 50 mM glycine at pH 8.5. After vortexing and incubating the mixture in the dark for 5 min, the fluorescence emission was measured at 485 nm with an excitation at 450 nm using a chemiluminescence spectrophotometer LS-55B (PerkinElmer, U.S.A.). For ANS binding fluorescence, κ-casein at 40 μM was combined with 25 μM ANS in 10 mM sodium phosphate (pH 8.0) for 10 min at the various temperatures indicated. The binding fluorescence was monitored between 400 and 650 nm with an excitation wavelength of 350 nm. Preparation of κ-Casein Hydrogel. The κ-casein fibrils prepared in 10 mM sodium phosphate, pH 8.0, at 80 °C for 18 h were subjected to dialysis against distilled water at room temperature for 24 h. The fresh distilled water was changed twice every 4 h and then left overnight. To produce the chemical-entrapped κ-casein hydrogel, the dialysis of DTT-treated κ-casein fibrils was carried out in distilled

■

RESULTS Fibrillar Polymorphism of κ-Casein. One-dimensional (1-D) protein nanofibrils of κ-casein, destined to form 2-D and 3-D suprastructures, were prepared by taking advantage of the fibrillar polymorphism of κ-casein fibrillation. The amyloid fibril formation of κ-casein is distinctive from other amyloidogeneses as its fibrillation kinetics lacks the lag period, representing the seed formation, typically observed with other amyloidogenic proteins.44 Besides the prevalent notion of nucleation-dependent fibrillation, therefore, the fibrillation mechanism for κ-casein has been proposed in two different models.45,46 The protein has been suggested to form amyloid fibrils from an initial oligomeric state, also known as a micellar state, via either direct assembly of structurally altered oligomeric species45 or self-assembly of activated monomers slowly released from the initial oligomers.46 These two assembly processes would thus result in the amyloid fibrils with distinctive morphologies. In 2732

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738

Biomacromolecules

Article

Figure 1. (A) Fibrillation of κ-casein in the presence and absence of DTT. The fibrillation kinetics of κ-casein was monitored with thioflavin-T binding assay. κ-Casein (80 μM) was incubated in 10 mM sodium phosphate, pH 8.0, at 37 °C in the absence (○) and presence (●) of 20 mM DTT. (B) TEM images of the κ-casein fibrils obtained in the absence (left) and presence (right) of DTT. The κ-casein fibrils collected after 72 h of incubation were visualized with TEM following a uranyl acetate staining. Scale bars represent 200 nm. Insets show the fibrils in 2-fold magnification. (C) Dynamic light scattering (DLS) analyses of κ-casein with and without DTT. κ-Casein (80 μM) was incubated at room temperature for 30 min in the absence and presence of 20 mM DTT and then subjected to DLS analysis. (D) Schematic representation of the κ-casein fibrillation mechanisms with and without DTT.45,46

Figure 2. (A) Temperature-dependent fibrillation kinetics of κ-casein. In the presence of 20 mM DTT, κ-casein (80 μM) was incubated at 20, 40, 60, and 80 °C, and the fibrillations were monitored with thioflavin-T binding fluorescence. (B) TEM images of the κ-casein fibrils obtained at the various temperatures. The fibrils incubated at 20 and 40 °C were collected after 72 h of incubation, while the fibrils at 60 and 80 °C were collected after 18 h. The scale bars represent 500 nm. (C) Liquid crystal state of the κ-casein fibrils. Optical images were obtained under cross-polarized light.

mechanisms by favoring the monomer assembly over the oligomeric association. Disintegration of oligomers by DTT was evaluated with dynamic light scattering (DLS) measurement as the average size of 21.4 ± 9.6 nm for the native oligomeric κ-caseins decreased down to 6.7 ± 1.5 nm upon the DTT treatment (Figure 1C). It was the reduced state of κcasein that enhanced the fibrillation kinetics. It is, therefore, pertinent to consider that the monomer assembly is favored under the reducing condition, while the oligomeric assembly proceeds under the pristine condition, which results in the DTT-dependent fibrillar polymorphism producing either

fact, depending on the presence of dithiothreitol (DTT), the fibrillation kinetics of κ-casein differed significantly along with the final fibrillar structures (Figure 1). In the presence of DTT, the fibrillation of κ-casein was accelerated and the amount of fibrils finally formed was also increased, as monitored by thioflavin-T binding fluorescence (Figure 1A). Comparatively thin and straight amyloid fibrils were produced from κ-casein in the presence of DTT, while thicker fibrils with a rugged texture were obtained in the absence of DTT (Figure 1B). Because native κ-casein oligomers are stabilized via disulfide bonds,47 the DTT treatment would differentiate the two 2733

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738

Biomacromolecules

Article

smooth surface fibrils or curly fibrils with rugged surface textures, respectively (Figure 1D). Two-Dimensional Fibrillar Alignment. To increase the aspect ratio of fibrils and thus improve their application potentials, the DTT-induced κ-casein fibrillation was carried out at high temperature because heat was previously demonstrated to facilitate the fibrillation presumably by altering the native structure of κ-casein.48 As κ-casein was incubated at different temperatures (20, 40, 60, and 80 °C) in 10 mM sodium phosphate at pH 8.0 in the presence of DTT, the fibrillation was proportionally augmented, as evidenced from the increase in thioflavin-T binding fluorescence (Figure 2A). At the highest temperature of 80 °C, however, the fibrillation quickly reached the maximum within 2 h, which was followed by a gradual decrease in fluorescence intensity over 25 h. Due to the conformationally and kinetically facilitated monomer association at the high temperature, the accelerated fibrillation was responsible for the sudden rise of the thioflavin-T fluorescence. For the subsequent gradual decrease in the fluorescence, however, the dye binding appeared to be hindered as the fibrils accumulated and experienced a possible postfibrillar assembly. In fact, thioflavin-T binding fluorescence for the fibrils collected at 25 h time-point was nonlinear and showed a downward curvature as the amount of fibrils increased while the fluorescence of the fibrils obtained at 2 h point increased proportionally (Supporting Information, Figure S1). In addition, the resulting κ-casein precipitates obtained at 25 h were shown to exhibit bright apple-green birefringence upon congo red binding under a polarized microscope, indicating that the aggregates were indeed the precipitates of amyloid fibrils (Supporting Information, Figure S2). When the fibrils thus formed were examined with TEM after an extended period of incubation, the fibrils became elongated with average lengths of 13.1 ± 3.8, 160.5 ± 56.6, 325.3 ± 124.5, and 482.1 ± 194.2 nm at 20, 40, 60, and 80 °C, respectively (Figure 2B). More intriguingly, the longest fibrils obtained at 80 °C were confirmed to align into a highly packed 2-D array structure (Figure 2B), indicating that the postfibrillar assembly resulting from the elongated fibrils with high aspect ratio gave rise to a nematic liquid crystal state of amyloid fibrils of κcasein. When those fibrils prepared at various temperatures were examined under cross-polarized light, apparent light transmission became more evident for the fibrils obtained at the higher temperatures (Figure 2C). The data suggest that the fibrillar liquid crystalline state is readily prepared via enhanced lateral packing of the elongated fibrils obtained during the high temperature incubation. Three-Dimensional Fibrillar Matrix Formation: Hydrogel. Hydrogel comprised of κ-casein fibrils was prepared in 3-D fibrillar matrix via dialysis of the ordered 2-D arrays against distilled water for more than 20 h at room temperature (Figure 3A) by taking advantage of the intrinsic instability of nematic liquid crystal state as previously demonstrated.49 Intriguingly, the κ-casein hydrogel consists of a multilayered planar structure unlike other protein-based hydrogels made of collagen50 or αsynuclein24 that had a fibrillar meshwork containing micro- or nanoscale water-entrapped pores formed between individual fibrils. The interlamellar space instead would retain water for the gelation of κ-casein 2-D arrays (Figure 3B). A magnified FE-SEM image of the κ-casein hydrogel indicated that the fibrillar nature was still maintained in the lamellar structure following the gelation (Supporting Information, Figure S3). The gel formation started to occur after 4 h of water dialysis

Figure 3. (A) Optical image of the κ-casein hydrogel. The hydrogel was produced via dialysis of the κ-casein fibrils obtained with 20 mM DTT at 80 °C for 18 h of incubation against distilled water at room temperature for 24 h. (B) SEM image of the κ-casein hydrogel. The scale bar represents 100 μm. (C) SEM images of the κ-casein hydrogel before and after its disintegration. Disintegration was carried out by submerging the hydrogel in 10 mM sodium phosphate buffer (pH 8.0) for 6 h. The scale bars represent 20 μm.

and completed after 20 h. It was also noticed that stiffness as well as elasticity of the hydrogel increased as the dialysis continued. The hydrogel formation was observed only with the water dialysis in the absence of any sort of salts because the dialysis carried out with 10 mM sodium phosphate at pH 6−8 failed to produce the hydrogel. This observation suggests that disruption of electrostatic interactions initially stabilizing the fibrillar alignment might be responsible for the gel formation during the water dialysis. Alternatively, the gel prepared against water was selectively disintegrated in 10 mM sodium phosphate at pH 8.0. Although the hydrogel was stable in water for more than 2 months, the gel readily dissolved in the phosphate buffer within 4 h. A more detailed structure, examined with SEM, indicates that the multilayered structure was fragmented into smaller pieces (Figure 3C). This salt-dependent disintegration phenomenon of κ-casein hydrogel, therefore, has been employed to improve its own application value as the gel has been utilized in controlled release of bioactive compounds. Controlled Release from the κ-Casein Hydrogel. To elaborate the controlled release, characteristic temporal and spatial release behaviors of fluorescein isothiocyanate (FITC) previously entrapped in the κ-casein hydrogel were examined with the diffusion-based gradient generator. The time-dependent FITC release was monitored as the gel was disintegrated in 10 mM sodium phosphate, pH 8.0. The fluorescence of escaping FITC was detected at a fixed position of the gradient channel, as indicated in the inset of Figure 4A. After immersing the FITC-gel prepared with 80 μM κ-casein into the phosphate solution, a lag period of 4 h was needed to observe the fluorescence, which followed a linear increase in the fluorescence for an additional 4−5 h until the intensity reached a plateau, while no fluorescence was detected for the FITCpreloaded gel placed in distilled water (Figure 4A). It was also found that the lag phase was completely disappeared when the gel was prepared with 40 μM κ-casein (Supporting Information, Figure S4). These facts indicate that the FITC-release was accompanied with actual disintegration of the hydrogel and the 2734

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738

Biomacromolecules

Article

Figure 4. (A) Time-dependent FITC release from the κ-casein hydrogel. FITC (50 μM) was incorporated in the κ-casein hydrogel (80 μM) during the dialysis process and the FITC-embedded hydrogel was placed in 10 mM sodium phosphate at pH 8.0. Fluorescence intensity was measured at 525 nm with an excitation wavelength of 460 nm. The representative scheme of the PDMS channel is provided, which was developed to analyze the temporal and spatial release of the FITC fluorescence (inset). (B) Spatial distribution of FITC released from the κ-casein hydorgel. The FITCembedded κ-casein hydrogel was immersed in the 10 mM phosphate buffer at pH 8.0 and its release was monitored with the fluorescence in the linear region of 600 μm indicated in the PDMS channel (inset).

Figure 5. (A) Optical image of the κ-casein hydrogel located in a 9 cm culture dish. The dish was sectioned into four parts and label separated as shown. (B) Average neurite lengths in the separate sections of the culture dish after 20 h of additional incubation with either free or entrapped retinoic acid within the hydrogel. The average neurite length was measured with the microscopic images of 100 different neurites for three separate sets each. The lengths are compared in percentage with a control set of neurons cultured in the absence of retinoic acid. (C) Optical microscopic images of SH-SY5Y cells treated with free RA (10 μM) and the RA-incorporated κ-casein hydrogel. Retinoic acid (10 μM) was entrapped in the κcasein hydrogel (160 μM) during the dialysis process. The gel was placed at the corner of “I” shown in panel (A) and incubated for 20 h. The 2-fold magnified images of the neurons and the neuritis in red are also presented at the bottom of each original image. The scale bars represent 100 μm.

when the fluorescence emerged from the gel at the highest level (Figure 4A). As the disintegration of gel and subsequent FITC release proceeded, the slope gradually decreased to the initial level of 0.022 (AU/μm). These data indicate that the spatial gradient of a compound released from the temporally controllable κ-casein hydrogel would influence its neighboring chemical targets in a spatially selective way. Controlled Release of Retinoic Acid and Its Influence over the Neuronal Differentiation. The κ-casein hydrogel was employed to demonstrate its use for temporal and spatial control over the neuronal differentiation. As a widely known neuronal differentiation factor, retinoic acid (RA) was

release thus could be temporally controlled by using different concentrations of κ-casein for the gel preparation. Spatial gradient of the escaping FITC was also evaluated at every 1 h point during the release with the designed gradient channel in the defined region between 0 and 600 μm as the FITC-preloaded hydrogel made of 80 μM κ-casein was disintegrated in 10 mM sodium phosphate, pH 8.0 (Figure 4B, inset). The spatial fluorescence gradient started to appear from the 5 h incubation point with a negative slope of 0.034 (AU/μm), while no prominent fluorescence gradient was detected beforehand. The negative slope became steeper and reached its maximum of 0.072 (AU/μm) at 6 h (Figure 4B) 2735

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738

Biomacromolecules

Article

assembly producing the distinctive fibrillar end product of unique surface structure of straight fibrils. The 1-D fibrillar structures in turn determine their subsequent 2-D and 3-D suprastructure formation. In this respect, the DTT-dependent assembly of κ-casein yielding the smooth-surfaced straight fibrils turns out to be crucial to elicit its application potential by producing the liquid crystal state and the hydrogel. It has also been found that the heat-dependent elongation of the straight fibrils of κ-casein resulting in the enhanced aspect ratio of individual fibrils has participated in the facilitated lateral association leading to the closely packed 2-D array. Concerning biocompatibility of κ-casein and diversity in chemical and physical modifications of protein-based materials, the temperature-promoted fibrillar alignment could be readily employed to organize functionalized biocompatible protein fibrils into the useful form of 2-D nematic liquid crystal state that could be applied in the development of various areas of chemical and biological sensors, power devices, catalytically reactive membranes, and tissue engineering. In fact, the liquid crystal formation depends on not only the physical property of the fibrillar aspect ratio but also the chemical nature of associating fibrils. As the Onsager theory51 suggests, the balanced electrostatic repulsive interaction between adjacent charged rod-shaped structures would be a critical part of the fibrillar alignment via the excluded volume interactions between fibrils. Because the hydrogel of κ-casein was generated from the well-ordered liquid crystal state by the desalting procedure of dialysis against distilled water, the optimal electrostatic interactions are believed to exist between the fibrils in the 2-D array. Alternatively, disruption of the optimal interactions would be responsible for the hydrogel formation comprised of the distinctive multilamellar internal structures. In other words, therefore, the hydrogel formation and its disintegration could be manipulated by altering the ionic environment. As a matter of fact, we were able to induce another state of hydrogel from a once-dissolved hydrogel in the phosphate buffer after redialysis against distilled water for 48 h (data not shown). This reversibility in the gel formation and disintegration indicates that the multilayered hydrogel structure and underlying interfibrillar association could be reversibly disrupted and restored by the addition and removal of salt ions, which would definitely augment the materialistic value of the hydrogel. In addition, we have also noticed that the gel formation was dependent upon the temperature of water dialysis because the gels were not produced at 4 °C but room temperature. The specific layer-by-layer interactions within the gel and the individual layer formation were presumed to be affected by the processing temperature. Controlled release of bioactive substances is critical to maximize the effectiveness of their biological activities. Because the κ-casein hydrogel has exhibited a delayed disintegration as represented by the lag phase of FITC-release, the matrix could be employed for the temporally and spatially controlled release of bioactive compounds. The temporal control could be achieved by modifying either the gel formation or its breakdown procedure. The lag period of the FITC release from the gel prepared with 80 μM κ-casein could be manipulated by altering the protein concentration for the gel formation, as noticed from the complete disappearance of the lag for the gel of 40 μM κ-casein (Supporting Information, Figure S4). Control of the lag period, therefore, could be effectively utilized to accomplish the temporally controlled release. By the same reason, diverse disintegration conditions

embedded within the κ-casein hydrogel and its selective influence on the differentiation of SH-SY5Y cells was examined. Following 24 h of preculture of the cells on a 9 cm dish, the RA-hydrogel made of 160 μM κ-casein was placed in section I as depicted in Figure 5A and the neuritic extensions were examined after 20 h of incubation in the designated sections of I−IV (Figure 5B,C). As a control, the cells were also cultured with unbound RA in solution at 10 μM. In the presence of RA, the neuronal growth was promoted by 80.3% from the cells cultured without RA during a fixed incubation period of 20 h (Figure 5B). As the neurons were cultured with the RAcontaining hydrogel, the neuronal extensions gradually decreased as the cells were located away from the gel. In the presence of the unbound RA, however, the cells were uniformly extended throughout the plate. The mean neurite lengths gradually decreased: 196.5, 161.8, 152.9, and 104.0% for sections I, II, III, and IV, respectively (Figure 5B,C). The neuronal outgrowth in section I was comparable with the neurons differentiated with the unbound RA, while that in section IV was almost the same level as the neuronal growth obtained in the absence of RA. This remarkable sectional difference in the neuritic growth clearly illustrates that the slowly released RA, which is expected to show a linear gradient, as demonstrated by the FITC-release in Figure 4B, has been translated into the discrete spatial gradient of neuritic extension as the neuronal differentiation factor was sequestered into the cells.

■

DISCUSSION As suggested in the previous studies,45,46 the DTT-dependent polymorphism of κ-casein amyloid fibrils has also been demonstrated in this study. It is apparent that the reduced κcasein has assembled differently from the κ-casein in the absence of DTT, presumably indicating that the monomeric association has been favored over the oligomeric unit assembly. κ-Casein has two cysteine residues (Cys-11 and Cys-88) that would contribute to inter- and intramolecular disulfide bond formation between κ-casein molecules. Upon the DTTtreatment, κ-casein has been shown to alter its structure by exposing a hydrophobic region(s), as confirmed with the increased ANS binding fluorescence (Supporting Information, Figure S5). This altered state of κ-casein presumably in the monomeric state, as demonstrated with the dynamic light scattering data (Figure 1C) could be responsible for the straight amyloid fibril formation (Figure 1B) which allows the fibrils to make close contact with each other. However, the surface roughness might not be the only condition for the fibrils to convert into the 2D-array since it also requires the enhanced aspect ratio of the elongated fibrils. Our result indicated that the fibrillar elongation has been achieved by the high temperature incubation at 80 °C. Because the temperature would exhibit dual effects on the κ-casein assembly by either altering protein conformation or stimulating collisional process, both effects could participate in the facilitated molecular assembly leading to the fibrillar extension. In fact, ANS binding study of κ-casein indicated that the protein showed significant structural alteration at the high temperatures of 60 and 80 °C (Supporting Information, Figure S6). In addition, kinetic stimulation of the monomeric protein appears to be also critical for the extension process to overcome the activation energy of the monomeric association. Based on the DTT-dependent fibrillar polymorphism, it is noticed that the initial state of κ-casein literally directs its fate of 2736

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738

Biomacromolecules

Article

Scheme 1. Schematic Representation of the Hierarchical Suprastucture Formation of κ-Casein Generating 1-D Protein Nanofibrils, 2-D Fibrillar Array of Liquid Crystal State, and 3-D Matrix of Hydrogel and Use of the Hydrogel to Control the Neuronal Differentiation

suprastructures derived from the biocompatible κ-caseins, therefore, are expected to be widely appreciated in various areas of future nanobiotechnology.

can also be harnessed, which include various salts and pH. In addition, the reversibility of the κ-casein hydrogel formation with the addition and depletion of salts could also contribute to improving its usefulness. Because the release from the κ-casein hydrogel was due to actual gel disintegration instead of chemical leakage, the reasonable amount of the entrapped substance like FITC was liberated in a rather short period of time, which resulted in a noticeable spatial gradient of the released compound unlike other gel systems.52,53 Although the spatial gradient once formed should be leveled off in solution as time goes by, the gradient could be imprinted into a spatial gradient of biological effects as the released substance encounters the cells. Because chemical effects to the cells have occurred through biological receptor interaction and subsequent uptake of the bioactive substance, a spatial chemical gradient would be reflected into the spatial cellular effects as demonstrated in this study of neuronal growth control. By allowing localization of a biological effect, therefore, a spatial control of the cellular activity could be readily exercised with the κ-casein hydrogel as a matrix for selective temporal and spatial controlled release, which could be eventually applied in the areas of drug delivery and tissue engineering. Before its actual use in vivo, however, possible toxic effects of the amyloid-based biomaterials should be carefully evaluated as they might be involved in various processes such as generation of reactive oxygen species,54,55 lipid membrane disruption,56,57 seed-dependent amplification of amyloid fibrils,58 toxic oligomeric intermediate formation,59,60 triggering of innate immune response mediated by macrophage,61 infectivity observed with prion protein,62,63 and apoptosis.64

■

ASSOCIATED CONTENT

S Supporting Information *

Six additional figures: Figure 1, Linearity of thioflavin T-binding fluorescence of the κ-casein fibrils; Figure 2, Congo red birefringence of κ-casein fibrils; Figure 3, Field-emission scanning electron microscope image of the κ-casein hydrogel; Figure 4, Disintegration rates of FITC-incorporated κ-casein hydrogels; Figure 5, Fluorescence intensity of ANS with κcasein in the presence and the absence of DTT; Figure 6, ANS binding fluorescence spectra of κ-casein obtained in the presence of DTT at various temperatures. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel.: +82-2-880-7402. Fax: +82-2888-1604. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2010-0009809). K.C. acknowledges the financial support from the National Research Foundation of Korea (NRF) funded by the Korea Ministry of Education, Science and Technology (MEST; The Creative Research Initiative Program for “Intelligent Hybrids Research Center”; No. 2010-0018290) and The WCU Program of Chemical Convergence for Energy and Environment (No. R31-10013).

■

CONCLUSION In this report, a single amyloidogenic protein of κ-casein has been engineered to create the hierarchical suprastructures (Scheme 1). Novel 1-D protein nanofibrils of κ-caseins were obtained with dithiothreitol by taking advantage of the mechanism-based fibrillar polymorphism of κ-casein amyloid fibrils. 2-D fibrillar arrays were produced as the nematic liquid crystalline state by controlling the fibrillar aspect ratio as the fibrils were elongated by incubating the κ-caseins at high temperature in the presence of dithiothreitol. Finally, 3-D matrix of a hydrogel was generated by controlling stability of the nematic liquid crystalline state with water-dialysis by reducing the balanced electrostatic interactions critical for maintaining the fibrillar alignment. The hydrogel was demonstrated to be used for the controlled release of bioactive compound like retinoic acid, which led to temporal and spatial control over the neuronal differentiation. These engineered

■

REFERENCES

(1) Lawrence, B. D.; Cronin-Golomb, M.; Georgakoudi, I.; Kaplan, D. L.; Omenetto, F. G. Biomacromolecules 2008, 9, 1214−1220. (2) Dong, J.; Sun, Q.; Wang, J.-Y.J.-Y. Biomaterials 2004, 25, 4691− 4697. (3) Wang, Y.; Kim, H.-J.; Vunjak-Novakovic, G.; Kaplan, D. L. Biomaterials 2006, 27, 6064−6082. (4) Lee, J.-H.; Kang, S.; Lee, S.-G.; Jin, J.-H.; Park, J. W.; Park, S. M.; Jung, S.; Paik, S. R. Acta Biomater. 2010, 6, 4689−4697. (5) Rochet, J. C.; Lansbury, P. T., Jr. Curr. Opin. Struct. Biol. 2000, 10, 60−68. (6) Harper, J. D.; Lansbury, P. T. Annu. Rev. Biochem. 1997, 66, 385− 407. 2737

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738

Biomacromolecules

Article

(7) Slotta, U.; Hess, S.; Spieß, K.; Stromer, T.; Serpell, L.; Scheibel, T. Macromol. Biosci. 2007, 7, 183−188. (8) Selkoe, D. J. Nature 2003, 426, 900−904. (9) Sacchettini, J. C.; Kelly, J. W. Nat. Rev. Drug Discovery 2002, 1, 267−275. (10) Hardy, J.; Selkoe, D. J. Science 2002, 297, 353−356. (11) Chiti, F.; Dobson, C. M. Annu. Rev. Biochem. 2006, 75, 333− 366. (12) Barnhart, M. M.; Chapman, M. R. Annu. Rev. Microbiol. 2006, 60, 131−147. (13) Costerton, J. W.; Stewart, P. S.; Greenberg, E. P. Science 1999, 284, 1318−1322. (14) Gebbink, M. F. B. G.; Claessen, D.; Bouma, B.; Dijkhuizen, L.; Wosten, H. A. B. Nat. Rev. Microbiol. 2005, 3, 333−341. (15) Talbot, N. J. Curr. Biol. 2003, 13, R696−R698. (16) Iconomidou, V. A.; Vriend, G.; Hamodrakas, S. J. FEBS Lett. 2000, 479, 141−145. (17) Iconomidou, V. A.; Chryssikos, G. D.; Gionis, V.; Galanis, A. S.; Cordopatis, P.; Hoenger, A.; Hamodrakas, S. J. J. Struct. Biol. 2006, 156, 480−488. (18) Fowler, D. M.; Koulov, A. V.; Alory-Jost, C.; Marks, M. S.; Balch, W. E.; Kelly, J. W. PLoS Biol. 2005, 4, e6. (19) Dobson, C. M. Methods 2004, 34, 4−14. (20) Fändrich, M.; Meinhardt, J.; Grigorieff, N. Prion 2009, 3, 89−93. (21) Kodali, R.; Wetzel, R. Curr. Opin. Struct. Biol. 2007, 17, 48−57. (22) Petkova, A. T.; Leapman, R. D.; Guo, Z.; Yau, W.-M.; Mattson, M. P.; Tycko, R. Science 2005, 307, 262−265. (23) Wetzel, R.; Shivaprasad, S.; Williams, A. D. Biochemistry 2007, 46, 1−10. (24) Bhak, G.; Lee, S.; Park, J. W.; Cho, S.; Paik, S. R. Biomaterials 2010, 31, 5986−5995. (25) Vetri, V.; Militello, V. Biophys. Chem. 2005, 113, 83−91. (26) Scheibel, T.; Parthasarathy, R.; Sawicki, G.; Lin, X.-M.; Jaeger, H.; Lindquist, S. L. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 4527−4532. (27) Fu, X.; Wang, Y.; Huang, L.; Sha, Y.; Gui, L.; Lai, L.; Tang, Y. Adv. Mater. 2003, 15, 902−906. (28) Pilkington, S. M.; Roberts, S. J.; Meade, S. J.; Gerrard, J. A. Biotechnol. Prog. 2010, 26, 93−100. (29) Corrigan, A. M.; Müller, C.; Krebs, M. R. H. J. Am. Chem. Soc. 2006, 128, 14740−14741. (30) Zhang, Y.; Gu, H.; Yang, Z.; Xu, B. J. Am. Chem. Soc. 2003, 125, 13680−13681. (31) Yang, Z.; Liang, G.; Wang, L.; Xu, B. J. Am. Chem. Soc. 2006, 128, 3038−3043. (32) Mahler, A.; Reches, M.; Rechter, M.; Cohen, S.; Gazit, E. Adv. Mater. 2006, 18, 1365−1370. (33) Silva, G. A.; Czeisler, C.; Niece, K. L.; Beniash, E.; Harrington, D. A.; Kessler, J. A.; Stupp, S. I. Science 2004, 303, 1352−1355. (34) Corrigan, A. M.; Müller, C.; Krebs, M. R. H. J. Am. Chem. Soc. 2006, 128, 14740−14741. (35) Hamley, I. W.; Castelletto, V.; Moulton, C.; Myatt, D.; Siligardi, G.; Oliveira, C. L. P.; Pedersen, J. S.; Abutbul, I.; Danino, D. Macromol. Biosci. 2010, 10, 40−48. (36) Mahler, A.; Reches, M.; Rechter, M.; Cohen, S.; Gazit, E. Adv. Mater. 2006, 18, 1365−1370. (37) Yoo, P. J.; Nam, K. T.; Qi, J.; Lee, S.-K.; Park, J.; Belcher, A. M.; Hammond, P. T. Nat. Mater. 2006, 5, 234−240. (38) Aggeli, A.; Bell, M.; Boden, N.; Keen, J. N.; Knowles, P. F.; McLeish, T. C. B.; Pitkeathly, M.; Radford, S. E. Nature 1997, 386, 259−262. (39) Gazit, E. FEBS J. 2007, 274, 317−322. (40) Gras, S. L.; Tickler, A. K.; Squires, A. M.; Devlin, G. L.; Horton, M. A.; Dobson, C. M.; MacPhee, C. E. Biomaterials 2008, 29, 1553− 1562. (41) Olsen, D.; Yang, C.; Bodo, M.; Chang, R.; Leigh, S.; Baez, J.; Carmichael, D.; Perälä, M.; Hämäläinen, E. R.; Jarvinen, M.; Polarek, J. Adv. Drug Delivery Rev. 2003, 55, 1547−1567. (42) Hamedi, M.; Herland, A.; Karlsson, R. H.; Inganäs, O. Nano Lett. 2008, 8, 1736−1640.

(43) Thorn, D. C.; Meehan, S.; Sunde, M.; Rekas, A.; Gras, S. L.; MacPhee, C. E.; Dobson, C. M.; Wilson, M. R.; Carver, J. A. Biochemistry 2005, 44, 17027−17036. (44) Ecroyd, H.; Koudelka, T.; Thorn, D. C.; Williams, D. M.; Devlin, G.; Hoffmann, P.; Carver, J. A. J. Biol. Chem. 2008, 283, 9012− 9022. (45) Leonil, J.; Henry, G.; Jouanneau, D.; Delage, M.-M.; Forge, V.; Putaux, J.-L. J. Mol. Biol. 2008, 381, 1267−1280. (46) Ecroyd, H.; Thorn, D. C.; Liu, Y.; Carver, J. A. Biochem. J. 2010, 429, 251−260. (47) Rasmussen, L. K.; HØjrup, P.; Petersen, T. E. Eur. J. Biochem. 1992, 207, 215−222. (48) Groves, M. L.; Wickham, E. D.; Farrell, H. M. J. Protein Chem. 1998, 17, 73−84. (49) Wong, I. Y.; Footer, M. J.; Melosh, N. A. J. Am. Chem. Soc. 2008, 130, 7908−7915. (50) Krewson, C. E.; Chung, S. W.; Dai, W.; Saltzman, W. M. Biotechnol. Bioeng. 1994, 43, 555−562. (51) Onsager, L. Ann. N.Y. Acad. Sci. 1949, 51, 627−659. (52) Hennink, W. E.; Jong, S. J. D.; Bos, G. W.; Veldhuis, T. F. J.; Nostrum, C. F. V. Int. J. Pharmacol. 2004, 277, 99−104. (53) Licciardi, M.; Grassi, M.; Stefano, M. D.; Feruglio, L.; Giuliani, G.; Valenti, S.; Cappelli, A.; Giammona, G. Int. J. Pharmacol. 2010, 390, 183−190. (54) Huang, X.; Atwood, C. S.; Hartshorn, M. A.; Multhaup, G.; Goldstein, L. E.; Scarpa, R. C.; Cuajungco, M. P.; Gray, D. N.; Lim, J.; Moir, R. D.; Tanzi, R. E.; Bush, A. I. Biochemistry 1999, 38, 7609− 7616. (55) Junn, E.; Mouradian, M. M. Neurosci. Lett. 2002, 320, 146−150. (56) Lashuel, H. A.; Hartley, D.; Petre, B. M.; Walz, T.; Lansbury, P. T. Nature 2002, 418, 291. (57) Caughey, B.; Lansbury, P. T. Annu. Rev. Neurosci. 2003, 26, 267−298. (58) Yamaguchi, K.; Takahashi, S.; Kawai, T.; Naiki, H.; Goto, Y. J. Mol. Biol. 2005, 352, 952−960. (59) Winner, B.; Jappelli, R.; Maji, S. K.; Desplats, P. A.; Boyer, L.; Aigner, S.; Hetzer, C.; Loher, T.; Vilar, M.; Campioni, S.; Tzitzilonis, C.; Soragni, A.; Jessberger, S.; Mira, H.; Consiglio, A.; Pham, E.; Masliah, E.; Gage, F. H.; Riek, R. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 4194−4199. (60) Outeiro, T. F.; Putcha, P.; Tetzlaff, J. E.; Spoelgen, R.; Koker, M.; Carvalho, F.; Hyman, B. T.; McLean, P. J. Proc. Natl. Acad. Sci. U.S.A. 2008, 3, e1867. (61) Klegeris, A.; Walker, D. G.; Mcgeer, P. L. Biochem. Biophys. Res. Commun. 1994, 199, 984−991. (62) Scott, M. R.; Will, R.; Ironside, J.; Nguyen, H. B.; Tremblay, P.; DeArmond, S. J.; Prusiner, S. B. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 15137−15142. (63) Holmes, T. C. Trends Biotechnol. 2002, 20, 16−21. (64) Hakansson, A.; Zhivotovsky, B.; Orrenius, S.; Sabharwal, H.; Svanborg, C. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 8064−8068.

2738

dx.doi.org/10.1021/bm300692k | Biomacromolecules 2012, 13, 2731−2738