Magnetically Actuated Heterogeneous Microcapsule-Robot for the

Jul 3, 2019 - Interfaces2019XXXXXXXXXX-XXX ...... Nguyen, D. K.; Son, Y. M.; Lee, N. E. Hydrogel Encapsulation of Cells in Core-Shell Microcapsules fo...
0 downloads 0 Views 10MB Size
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Magnetically Actuated Heterogeneous Microcapsule-Robot for the Construction of 3D Bioartificial Architectures Yang Liu,† Gen Li,‡,§ Haojian Lu,† Yuanyuan Yang,† Zeyang Liu,∥ Wanfeng Shang,*,‡ and Yajing Shen*,† †

Department of Biomedical Engineering, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong 999077, China Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China § Beijing Institute of Technology, Beijing 100081, China ∥ Stem Cell Therapy and Regenerative Medicine Lab, Tsinghua-Berkeley Shenzhen Institute, Shenzhen 518055, China

Downloaded via UNIV OF SOUTHERN INDIANA on July 24, 2019 at 15:56:47 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



S Supporting Information *

ABSTRACT: Core−shell microcapsules as one type of the most attractive carriers and reactors have been widely applied in the fields of drug screening and tissue engineering owing to their excellent biocompatibility and semi-permeability. Yet, the spatial organization of microcapsules with specific shapes into threedimensional (3D) ordered architectures still remains a big challenge. Here, we present a method to assemble shapecontrollable core−shell microcapsules using an untethered magnetic microcapsule-robot. The microcapsule-robot with the shape-matching design can grab the building components tightly during the transportation and assembly processes. The core− shell feature of the microcapsule effectively prevents the magnetic nanoparticles from interacting with bioactive materials. The assembly results of cell-loaded heterogeneous microcapsules reveal that this strategy not only allows the magnetic microcapsule-robot to work in different workspaces in vitro for the creation of 3D constructions but also offers a noninvasive and dynamical manipulation platform by remotely controlling the position and orientation of the soft and liquid-like microcapsule components individually. KEYWORDS: magnetic microcapsule-robot, heterogeneous, untethered, 3D bioartificial architectures, on-demand 3D assembly



INTRODUCTION Microcapsule is a kind of artificial scaffold with a bio-friendly aqueous core and a semipermeable shell, which can not only provide a mimic native three-dimensional (3D) tissue culture microenvironment in vitro but also separate encapsulated materials from the outer environment and exchange matters.1−3 Owing to these advantages, the core−shell microcapsule has been widely used as carriers and reactors for drug delivery, cell culture, and tissue regeneration in biomedical and tissue engineering.4−9 To reveal the structure− function relationship and the underlying mechanisms of biological process, one critical requirement is to mimic native 3D tissue architectures consisting of specific structural and functional units with a defined spatial distribution.10−12 Therefore, for the construction of complex 3D microcapsule tissue-like architectures, two indispensable factors, i.e., 3D heterogeneous microstructure fabrication and controllable ordered-assembly, are essentially required. At the current stage, although the construction of heterogeneous microcapsules, improving traditional approaches with respect to monotonous spherical and fibrous ones, has recently been reported,7,8,13−15 the spatial organization of heterogeneous © 2019 American Chemical Society

core−shell microcapsules in a controllable manner is still of great challenge. Self-assembly can realize self-alignment of smaller repeating monomer units into a larger-size complex tissue construction by molecular recognition or suitable driving forces (i.e., capillary and electrostatic forces). Although it is a parallel and fast assembly strategy, the probabilistic nature of selfassembly often results in the nondirected assembly.16−24 To increase the assembly precision and yield, the guided assembly based on microfluidic and mechanical manipulation methods has been developed and widely applied in the assembly of heterogeneous tissue-like artificial architecture.25−29 However, these technologies have still been plagued by a few shortcomings. Due to the direct assembly via the velocity and direction of flow, the microfluidic-based assembly is only adept at one-dimensional constructs.26,30,31 Additionally, though the mechanical manipulation strategy provides controllable manipulation in the position and orientation for the Received: March 28, 2019 Accepted: July 3, 2019 Published: July 3, 2019 25664

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces

Figure 1. Schematic illustration of the 3D heterogeneous ACA microcapsule fabrication and magnetic-directed assembly system. (a) Visible-lightinduced electrodeposition chip based on TiOPc was developed and adopted to fabricate the 2D Ca-alginate hydrogel template. The 3D heterogeneous ACA microcapsule with a defined structure was obtained using the 2D Ca-alginate hydrogel as the sacrifice template. (b) Microcapsule-encapsulating MNPs play the role of microcapsule-robots to assemble and locomote the microcapsule building blocks using gradientbased magnetic actuation on the C10F18−H2O interface.

invasive, biocompatible, guided, and on-demand 3D assembly strategy is compelling to meet the urgent needs in tissue engineering. In this work, we present a broadly applicable method that allows the fabrication of a 3D magnetic microcapsule-robot with tunable size and morphology, and spatially directed assembly using remote magnetic actuation. The core−shell alginate−chitosan−alginate (ACA) microcapsule-encapsulated MNPs plays the role of magnetic microcapsule-robots (MMcrobots). The semipermeable ACA layer separates bioactive materials from MNPs, thereby no requirements for releasing of MNPs from cell-loaded hydrogel microstructures. Gradientbased magnetic actuation is adopted in this manuscript due to the laconic, precise, and controllable merits. Through the precise manipulation of the MMc-robot, several microcapsule building blocks are successfully assembled, locomoted, and posture controlled at the interface between water and perfluorodecalin (C10F18). In addition, benefiting from the flexible creation of 3D ACA microcapsules with diverse morphologies, the MMc-robot can be fabricated according to the structural sizes of a microcapsule component for the shapematching effect, which can grab the building units tightly during the transport and assembly processes. Particularly, the noninvasive manipulation avoids the destruction of these soft and liquid-like microcapsule architectures. Furthermore, the flexibly and remotely manipulating method not only allows the micro-robot to work in open environments, but also has the potential to work in enclosed spaces such as the human body,

construction of complex tissue-like architectures, the intractable issue that end effectors are often connected with an actuating device restricts this strategy to only validate in open working environments.28,32 Moreover, the problem regarding the disruption of the fragile hydrogel caused by a rigid manipulator still remains to be unresolved. To overcome these challenges, noninvasive fields have been applied to assemble fragile hydrogel microstructures for complex 3D artificial constructs. For instance, an optically induced dielectrophoretic (ODEP) strategy can realize the rapid assembly of microgels,33−35 because the virtual electrode enables building the unit to be selected and captured individually.36 However, the position and orientation of components will be inevitably disturbed during the process of virtual electrode transported “target object” due to the nonselectivity of the polarized force of ODEP. By contrast, the magnetically driven robot assembly platform provides an inexpensive, versatile, and dynamic manipulation method, which can control the position and orientation of an individual microgel without any chemical fuels and physical contacts of mechanical operations. In particular, magnetic nanoparticles (MNPs) loaded in cell-encapsulating hydrogels (M-gels) have been commonly used as magnetic micro-robots for the remote microgel manipulation.37,38 Even though this strategy opens new venues for the construction of tissue-like hydrogel artificial architectures, the release of MNPs from these biomimetic composite materials still needs to be proven due to the risk of potential cytotoxicity of heavy metal.39 Therefore, a non25665

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces

Figure 2. 3D ACA microcapsule fabrication mechanism. Three steps were carried out to obtain 3D ACA microcapsules. Step 1, Ca2+ and alginate crosslinking by the light-induced electrodeposition device; step 2, the obtained 2D Ca-alginate hydrogel sheets are sequentially encapsulated by the chitosan and alginate coatings, and then the alginate−chitosan−alginate (ACA) shell is formed on the surface of the 2D hydrogel structure; step 3, after treatment of the ACA-coated Ca-alginate structure with sodium citrate, the solid-like cores liquefy due to the replacement of Ca2+ with Na+.

realize the locomotion and manipulation. Moreover, the actuation force is limited by the step-out frequency as the rotating magnetic field frequency increases. Consequently, the drag force generated by the gradient magnetic field is more suitable for the assembly of multi-building blocks and the posture control. The light-induced electrodeposition device consists of a top indium tin oxide (ITO) glass (cathode), a working chamber, and a bottom ITO substrate coating a thin photoconductive layer of TiOPc (anode). The exposed TiOPc film exhibits high photoconductivity when an optical pattern as a virtual electrode is projected onto the TiOPc plate (see more details in the Supporting Information).33,40,41 A PC controller is used to generate the programmable optical pattern, and the illumination source was projected by a projector. The optical pattern was scaled down by an objective lens (10×, Olympus Inc.). A direct current voltage was supplied by a signal generator (2601B, Keithley) and applied between the two electrodes. The deposition solution was introduced into the working chamber, and the uniform electric field in the liquid is generated by means of an externally applied voltage to trigger the Ca-alginate crosslinking when the optical pattern projected on the photoconductive chip. The programmable 2D Caalginate hydrogel template can be easily created because of the flexible virtual electrodes. Herein, a four degrees of freedom (4-DOF) electromagnetic coil system is developed (Figure S2). The manipulation of the MMc-robot was conducted in a small well surrounding by two sets of orthogonal electromagnetic coils that provide a magnetic field with the magnetic flux densities (Fe core) up to 34 mT. The sample dish is placed in the middle of the small well, and the assembly and locomotion of the 3D microcapsule building units were recorded by a home-made upright microscope. Formation of 3D Heterogeneous ACA Microcapsules. Three steps were carried out to obtain 3D heterogeneous ACA microcapsules, which includes: (1) Ca-alginate hydrogel

which holds great promise for biomedical and tissue engineering.



RESULTS AND DISCUSSION 3D Heterogeneous ACA Microcapsule Fabrication and Magnetic-Directed Assembly System. The successful construction of the 3D heterogeneous liquid-like ACA microcapsule was performed using the programmable microcapsule fabrication device (Figures 1a and S1). Briefly, twodimensional (2D) Ca-alginate hydrogel microstructures were fabricated by a TiOPc-based visible-light-induced electrodeposition chip. The electrochemically induced release of the solubilized calcium ion (Ca2+) from insoluble calcium carbonate (CaCO3) and the occurrence of the ionic crosslinking of Ca2+ and alginate chain occurred only in the projected region. After the deposition, the TiOPc plate was immersed into a petri dish to flush away the uncrosslinked deposition solution and collected the crosslinked Ca-alginate hydrogel microstructures. Then, the crosslinking 2D hydrogels ware transferred into a centrifuge tube filled with chitosan solution, which allows chitosan to react with Ca-alginate microstructures. The tube was shaken gently to prevent the microstructures from sticking together. The chitosan-coated Ca-alginate gel structures were washed and transferred into another tube filled with alginate solution and the alginate coating as well as the chitosan coating process were carried out. The coated gel structures were treated with sodium citrate solution to obtain the 3D ACA microcapsules. The obtained 3D microcapsule building blocks and MMc-robots were transferred to the gradient-based magnetic-directed assembly system for manipulation (Figure 1b). Compared with the actuation force generated by the magnetic torque, the gradientbased magnetic actuation has no requirement for the specific structural design of the actuator, e.g., length helical tail. The magnetic torque is utilized to rotate the asymmetric structure, which generates the complex flow field propulsion force to 25666

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces

Figure 3. Characterization of 2D Ca-alginate and 3D microcapsule. (a) Time dependence of deposition current at four different applied voltages (3, 4, 5, and 6 V). The larger applied voltage would induce a faster deposition rate. In this experiment, 3 V is the threshold voltage for electrodeposition, and the deposition current reaches the peak value after 70 s deposition period. (b) Electrodeposition heights of Ca-alginate hydrogel induced by the applied voltage of 4, 5, and 6 V for 70 s were recorded by a confocal laser scanning microscope. Scale bars are 200 μm. (c) Relationship between the polyelectrolyte complexation reaction period and the microcapsule membrane thickness. The thickness of the microcapsule wall increases with the increase of reaction time. (d) Thicknesses of AC microcapsules at four different reaction times (2, 6, 12, and 20 min) were measured by the pixel-calculating method. Scale bars are 5 μm.

Figure 4. 3D heterologous ACA microcapsule building blocks were fabricated on-demand. (a-i, b-i, c-i) 2D Ca-alginate hydrogel templates with different shapes of clover-, gear-, and three arms were fabricated using the programming electrodeposition system. (a-ii, b-ii, c-ii) corresponding 3D ACA microcapsules. (a-iii, b-iii, c-iii) Side view images of these microcapsules. The swelling degree is decreased with the increase of the encapsulation process time. (a-iv, b-iv, c-iv) Fluorescence images of these 3D microcapsules. Scale bars are 400 μm.

template fabrication was performed using the light-induced electrodeposition system; (2) chitosan and alginate were coated on the surface of the Ca-alginate hydrogel template to form the ACA shell; (3) the ACA-coated ca-alginate hydrogel template was treated with sodium citrate to carry out the liquefaction process by the ion exchange reaction sequentially

to achieve the liquid-like ACA microcapsule. The SEM images of core−shell microcapsule are shown in Figure S3. More details related to the 3D microcapsule fabrication are described in the Supporting Information and our previous works.13,42,43 The schematic illustration of the chemical reaction mechanism 25667

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces

Figure 5. Untethered magnetic-directed manipulation and assembly of lock-and-key-shaped 3D heterogeneous microcapsules. (a-i) A 4-DOFs electromagnetic coil system was developed and used to arrange the microcapsule building blocks into the shape-customized construction. The microcapsule building units were randomly placed on the interface of C10F18−H2O and assembled by the MMc-robot. (a-ii, a-iii) MMc-robot was dragged to the local maximum magnetic field due to the magnetic field gradient and aligned along the magnetic induction line direction under the magnetic torque. (b(i−v)) Time sequential microscope images show the untethered magnetic-directed assembly of 3D ACA microcapsule building blocks. The microcapsule building units and MMc-robot are outlined by the colored and black dotted line, respectively. The white arrow indicates the shadow of the MMc-robot. (c(i−iii)) Posture adjustment of the lock-and-key microcapsule structure on the interface of C10F18−H2O. (c-iv) Trajectory of the MMc-robot approaches, connects, and transports the lock-and-key microcapsule architecture. Scale bars are 400 μm.

electrodeposition, which indicates that the deposition rate of Ca-alginate hydrogel begins to decline. The deposition process can be observed and judged clearly by monitoring the change of current. Under different conditions of the applied voltages of 4, 5, and 6 V, the heights of hydrogel after 70 s deposition have been recorded by a confocal laser scanning microscope, respectively (Figure 3b). The constant applied voltage of 5−6 V was adopted in this experiment, which can form the firm Caalginate hydrogel microstructures in a short deposition period (about 30 s). Figure 4a−c illustrates the on-demand 3D heterologous ACA microcapsules fabrication using the light-induced electrodeposition device. 2D Ca-alginate hydrogel templates with various different shapes such as clover, gear, and three arms were formed in a controllable manner according to the light pattern. These 2D hydrogel templates were almost opaque because the CaCO3 nanoparticles are not completely dissolved by H+ during the electrodeposition process, as shown in Figure 4a-i,b-i,c-i.44 After chitosan and alginate coating, these ACA microcapsules exhibited perfect shapes and sharp edges since the ACA membrane was formed on the surface of the 2D Caalginate hydrogel templates. The nearly transparent micro-

for 2D Ca-alginate crosslinking and 3D microcapsule liquefaction is shown in Figure 2. During the Ca-alginate hydrogel growth, the constant voltage mode was adopted. The proton (H+) near the anode increased in accordance with the deposition time, and the concentration of Ca2+ increased in sequence, thus the deposition current raised up rapidly. The conductivity was decreased gradually as the ionic crosslinking occurs between the Ca2+ and alginate chain, and the deposition current started reducing and then tended to became stable gradually. In this work, four different voltage values (3, 4, 5, and 6 V) were applied to investigate the relationship between the applied voltage and the deposition rate. The effect of applied voltage on hydrogel growth is beneficial for judging the electrodeposition state and choice of the optimized deposition parameters, as shown in Figure 3a. Under the current experimental conditions, the threshold voltage is 3 V to trigger the electrochemical decomposition of water. The larger applied voltage would induce a faster Ca2+ release, thereby resulting in a short deposition time requirement. When the applied voltage increased to 6 V, the deposition current reached to the peak value after 70 s (the red dotted line in Figure 3a) 25668

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces

Figure 6. Heterogeneous 3D cell-encapsulating microcapsule assembly. (a1) A 3D hexagon-shaped ACA microcapsule was fabricated as a foundation architecture according to the designed optical pattern (a). (b−d) Series of evolving structures were designed based on the hexagonshaped pattern by increasing the number of the heads and tails. (b1−d1) Optical images of the corresponding 3D ACA microcapsule architectures. (c2) Schematic illustration of microcapsule building blocks arranged in a line. (c3) Optical image of a part of the ordered biological microcapsule architectures manipulated by the untethered magnetic microcapsule-robot. (c4) Corresponding fluorescence image. (d2) Schematic illustration of microcapsule building blocks assembled into a tissue-like network. (d3, d4) Optical and fluorescence images of the part of the tissue-like microcapsule network, respectively. Scale bars are 400 μm.

3D Microcapsule Assembly by Untethered Magnetic Actuated MMc-Robot. The proposed remotely actuated and controlled magnetic microcapsule-robot strategy has the advantage of the dynamic orientation control for the soft and liquid-like microcapsules assembly. The 3D ACA microcapsules with the shape-matching design were first fabricated by the programmable dimension ascending process and then precisely arranged into a “lock-and-key” geometry using the MMc-robot coding at the C10F18−H2O interface. The microcapsules can be either fabricated one by one or pattered in an array before being transported into the assembly dish for further manipulation. To code building units into 3D ordered architecture, an untethered MMc-robot was produced by the proposed heterogeneous microcapsule fabrication method. The magnetic robot is composed of MNPs encapsulated in an ACA microcapsule and it is magnetized in a uniform magnetic field with 1 T magnetic intensity, which is generated via two permanent magnets. The direction of magnetization is along the arrow direction of the MMc-robot’s central axis. After magnetization, the MMc-robot and microcapsule building blocks were transferred to an assembly dish and placed in the electromagnetic coil system with four electromagnets surrounding the workspace (Figure 5a-i). The microcapsule building blocks were randomly placed on the interface of C10F18−H2O and manipulated by the MMc-robot controlled

capsules free of CaCO3 nanoparticles are replaced by the ion exchange reaction of Ca2+ with Na+ (Figure 4a-ii,b-ii,c-ii). Note that the thickness of the microcapsule membrane not only relates to the permeability of microcapsule but also affects the swelling degree of the microcapsule. The thinner membrane is more suitable to supply nutrients and oxygen for biomaterials.45,46 As the encapsulation process time increases, the deepened interaction between the chitosan module and the alginate chain leads to thickening and strengthening of the microcapsule membrane, as shown in Figure 3c. Therefore, the swelling degree of the microcapsule decreases as the thickness of the microcapsule membrane increases (Figure S4, Video S1). The thicknesses of microcapsule walls were measured by the pixel-calculating method at four different reaction times of 2, 6, 12, and 20 min, respectively (Figure 3d). The side view images of these 3D microcapsules demonstrate the different swelling statuses of microcapsule with three different membrane thicknesses, 9, 15, and 25 μm in sequence, as shown in Figure 4a-iii,b-iii,c-iii. The fluorescence images of these 3D microcapsules are shown in Figure 4a-iv,b-iv,c-iv. In addition, the molecular weight, crosslinking time, concentration, and pH of chitosan and alginate can also affect the thickness of the microcapsule membrane (Figure S5). In this experiment, the layer thickness of the ACA microcapsules is in the range of 8−15 μm. 25669

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces

demonstrated in this work, as shown in Figure S6. A hexagonshaped pattern with a center hole was designed as a fundamental architecture (Figure 6a). The corresponding 3D microcapsule is shown in Figure 6a1. Based on this pattern, we increased the number of the “heads” and “tails” on the sides of hexagonal pattern gradually (Figure 6b−d), and a series of evolving 3D microcapsules with the defined structures were fabricated flexibly (Figure 6b1−d1). In this experiment, we performed the construction of ordered biological architectures by assembling the cell-loaded 3D microcapsule building units. Figure 6c3 shows that the biological microcapsule building units were aligned by the presented untethered MMc-robot manipulation method. The corresponding fluorescence image is shown in Figure 6c4. Moreover, the proposed manipulation method not only allows arranging the 3D heterogeneous microcapsule architectures in a line (Figure 6c2) but also can form the tissue-like network by the dynamically controllable magnetic microcapsule-robot manipulation (Figure 6d2). The advantage is beneficial to the cellomics and system biology studies. The optical and fluorescence images of the partial tissue-like network are shown in Figure 6d3,d4, respectively. The on-demand ACA microcapsule fabrication method enables the MMc-robot to be produced according to the structural sizes of microcapsule building blocks. The shapematching design is conducive to the transport and assembly processes. In addition, our untethered magnetic microcapsulerobot coding approach offers an effective assembly strategy over the soft and liquid-like microcapsule tissue architectures. As shown in Figure 6c3,d3, it is found that the excellent alignment between the microcapsule components brings about the perfectly matched architecture. Sometimes, the assembly of 3D heterogeneous microcapsule architectures would not proceed perfectly. The imperfect assembly could be attributed to the following points. First, the perfect assembly requires the microcapsule components to be fabricated with well-designed shape and size, so that the microcapsule building blocks can align and engage well with each other. There is a main challenge that the microcapsule dimensions have to be changed after the treatment of chitosan coating and sodium citrate liquefaction. Moreover, the 2D Ca-alginate hydrogel template affects the morphology of the microcapsule directly. Tuning the shape and size of the 2D hydrogel template by adjusting the parameters of light-induced electrodeposition, such as the electrodeposition period and applied voltage, is very important for further achieving highly matched microcapsule assembly. To overcome these challenges, the production process and experimental parameters will be optimized in our future work. Second, due to the MMcrobot coding at the C10F18−H2O interface, the microcapsule components remain disturbed by the liquid flow of the upper layer aqueous solution during the manipulating process, which will increase the difficulty of MMc-robot alignment, especially for some microcapsule architectures that are not very well fitted with each other. Therefore, reducing the disturbance of liquid flow to retain microcapsule components at a fixed or desired position is necessary.

through the algorithms of dynamically regulating the magnetic field in response to users’ command. Magnetic forces are imposed directly to the MMc-robot for the translation and orientation control (Figure 5a-ii,iii). The nearly transparent microcapsule building blocks are outlined by the colored dotted line (marked as the green arrows), and the microcapsule robot is outlined by the black dotted line (marked as the black arrow). The MNPs encapsulated in the capsule robot makes it opaque compared with the other building blocks. The shadow of the MMc-robot can be observed clearly (marked as the white arrow), as shown in Figure 5b-i. Figure 5b(i−v) represents a continuous sequence of 3D microcapsule architecture units’ assembly process by an untethered MMc-robot. One of the architecture’s convex part was fitted to another concave part. The four 3D soft and liquid-like microcapsule building units were aligned into an ordered construction (Video S2), which demonstrates that the presented MMc-robot manipulation method is capable of dynamically adjusting the orientation and position of each building unit effectively and flexibly. It allows the shapeselective lock-and-key geometry to be formed from more than one component. In addition, to illustrate the capability of the MMc-robot in handling the artificial architectures regardless of the size and weight, the assembled lock-and-key structure as a whole object was manipulated by the 3D MMc-robot (Figure 5c). The magnetic torque provided by the external magnetic field can be utilized to regulate the posture of the MMc-robot. The snapshots of the lock-and-key microcapsule architecture locomotion processes are shown in Figure 5c(i−iii) and captured from Video S3. The results show that the MMc-robot pushed the ordered construction rotating almost 180° at the C10F18−H2O interface. Next, the architecture was locomoted continuously for creating a horizontal posture. The MMcrobot posture regulation ensures each 3D ACA microcapsule building unit to assemble in a controllable manner so that the whole structure is able to remain stable and not easy to fall apart during the posture adjustment process. Moreover, it can also be observed that the manipulation speed has no significant variation in pushing the speed of the individual unit. Therefore, the push speed is not greatly affected by the microcapsule artificial construction size or weight. Time-lapse images in Figure 5c-iv show the whole motion trajectory that MMc-robot approaches, connects, and transports the lock-and-key microcapsule architecture. The MMc-robot did not need to be removed after the manipulation due to the isolation of the ACA microcapsule. These results show that the presented magnetically actuated microcapsule-robot coding method has great potential for the applications in the construction and repair of 3D ordered heterogeneous microcapsule architectures. Heterogeneous 3D Cell-Encapsulating Microcapsule Assembly. Constructing the cell-loaded microcapsule architectures with a defined 3D structure has a significant potential for tissue engineering, as it can fulfill a variety of important tasks, including personalized medicine testing and complex tissue reconstruction. Before an artificial tissue-like microcapsule architecture was constructed, the C2C12 myoblast cellencapsulated microcapsules with designed sizes and shapes were fabricated using the 3D ACA microcapsule fabrication system. The high viability and proliferation of cells seeded in ACA microcapsules with the liquid core have been demonstrated by Zhang et al.47 The high viability (>90%) of cells seeded in the liquid-like ACA microcapsules has also been



CONCLUSIONS In summary, we present a flexible and versatile method for 3D heterogeneous microcapsule fabrication and spatially directed assembly using untethered magnetically actuated microcapsulerobot. Various 3D on-demand microcapsules can be produced easily and quickly by implementing the programmable 3D 25670

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces

150i). The cells grown to 90% were detached from the Petri dish using 0.25% trypsin to obtain the cell suspension solution. Cell Fluorescence Staining and Characterization. Before microscopic observation, C2C12 myoblast cells were stained with fluorescent dyes. After cell centrifugation, C2C12 cells were resuspended in cell tracker green and red fluorescent dye, respectively. Then, the stained cells were cultured for 45 min in an incubator. After that, the stained cells were centrifugated again and the fluorescent dying solution was removed. Prior to the experiment, the cells were resuspended in the CaCO3-alginate mixed solution for the fabrication of cell-laden hydrogels. The bright-field images and fluorescence images of 3D Ca-alginate hydrogel microstructures and 3D ACA microcapsules were obtained using an Eclipse Ni microscope (Nikon, Tokyo, Japan). Confocal fluorescence images were obtained using a Leica TCS SP8 confocal laser scanning microscope (Leica, Germany). Videos were recorded by a home-made upright microscope. Home-Made Electromagnetic Coil System. The whole system was composed of a PC controller, data acquisition board (GTS-800PV-PCI-G, Googol Technology Limited), current driver (145391, MAXON Computer Limited), and microscopy imaging system for visual feedback. During the assembly, both the magnetic torque and magnetic drag force were applied to the MMc-robot in the external magnetic field. The MMc-robot was aligned along the magnetic induction line direction under the magnetic torque. The magnetic micro-robot was dragged to the local maximum magnetic field due to the magnetic field gradient. Therefore, during the experiment, the magnetic torque and drag force of the magnetic micro-robot generated by the external magnetic field are able to be calculated

microcapsule fabrication method. The bioactive materials and MNPs were encapsulated, respectively, into the 3D ACA microcapsules and they can be regarded as the assembled component and the soft magnetizable actuator, respectively. The semipermeable ACA layer provides an aqueous microenvironment that promotes the development of encapsulated cells and prevents bioactive materials from contacting with MNPs. The orientation and position of soft and liquid-like microcapsule building blocks can be adjusted by the MMcrobot dynamically at the C10F18−H2O interface. Furthermore, we demonstrated that the presented manipulation method can be applied to cell-loaded microcapsules without damage and toxicity. From the above, the proposed method of 3D heterogeneous microcapsule fabrication and assembly holds great promise to construct bioartificial architectures, which will offer new opportunities for promoting the development of tissue regeneration and drug screening applications.



EXPERIMENTAL SECTION

Materials. Sodium alginate powder (300−400 mPa s), methyl ethyl ketone (MEK), cyclohexanone, and poly(vinyl butyral) (PVB) powder were purchased from Acros Organics. Calcium carbonate nanoparticles (CaCO3, Φ 30−50 nm) were obtained from Haofu Chemistry Co. Ltd (China). Titanyl phthalocyanine (TiOPc) powder was acquired by Zhongmin Co. Ltd (Tianjin, China). Indium tin oxide glasses (ITO, 30 × 30 mm2, 15−30 Ω) were bought from the Taobao website. Chitosan (100−200 mPa s) and magnetite microspheres (M120286, Φ 300−400 nm) were obtained from Aladdin Bio-Chem Technology Co. Ltd (Shanghai, China). Sodium citrate tribasic dehydrate (S4641) was purchased from Sigma Aldrich Chemical Co. Ltd (St. Louis, MO). Fetal bovine serum (FBS), trypsin, penicillin, and streptomycin were acquired from Gibco (Carlsbad, CA). Cell tracker Green CMFDA dye and Red CMTPX dye were purchased from Life Technologies (Waltham, MA). C2C12 myoblasts cells and Dulbecco’s modified Eagle medium (DMEM) culture medium were provided by ATCC (Manassas, VA). Ultrapure water was obtained from Milli-Q Integral Water Purification System. Preparation of Deposition Solution. The deposition solution was prepared by suspending CaCO3 nanoparticles (0.5%) into a sodium alginate solution (1%) and sonicating for 10 min. Then, we performed magnetic stirring at 1200 rpm for 12 h to prepare the deposition solution. Chitosan and Sodium Citrate Solution Preparation. The chitosan solution was prepared by dissolving chitosan in saline at pH 5 by adding acetic acid. The pH of the solution was brought back to 6.5−6.6 using NaOH titration after the chitosan was dissolved completely. The sodium citrate was dissolved in 0.45% (w/v) NaCl solution to prepare 55 mM sodium citrate solution. Preparation of the Mixed Deposition-MNP Solution. The monodisperse magnetite microspheres with a carboxyl functional group (−COOH) solution were centrifuged at 3000 rpm for 2 min and then the solution was removed. After centrifugation, the MNPs were homogeneously resuspended in the deposition solution. TiOPc Solution Preparation. First, the MEK solution (5 mL) was added into a cyclohexanone solution (5 mL) and stirred at 700 rpm for 20 min to obtain the mixed solution. Next, TiOPc powder (1.5 g) and PVB powder (0.75 g) were added to the mixture solution followed by vigorous stirring (1200 rpm) for 10 h. Optoelectronic Chip Fabrication. One drop of TiOPc solution (500 μL) was dropped onto the surface of an ITO glass using a pipette. Then, the TiOPc solution was flattened using spin coater at a speed of 500 rpm for 15 s and 1200 rpm for 60 s, respectively. After that, the coated optoelectronic chip was baked at 120 °C for 30 min to harden the TiOPc layer. Cell Culture. C2C12 myoblast cells were cultured in a Petri dish containing DMEM supplemented with 10% FBS and 1% penicillin/ streptomycin. The culturing conditions were 37 °C and 5% CO2 in a CO2 constant-temperature incubator (Thermo Scientific Heracell

Tamr = Vamr·M × B(x , y , z)

(1)

Famr = Vamr· (M ·∇)·B(x , y , z)

(2)

where Tamr represents the magnetic torque of the actuation magnetic micro-robot, Vamr represents the volume of the actuation magnetic micro-robot, M represents the magnetization of the actuation magnetic micro-robot, B(x, y, z) represents the magnetic field induction intensity, Famr represents the magnetic drag force of the actuation magnetic micro-robot, and ▽ represents the magnetic gradient operator of the magnetic field. When it comes to position control of the 3D ACA magnetic microrobot, the magnetic drag force is starting to play a key role. Apart from the magnetic torque and magnetic drag force, the magnetic microrobot motion environment should also be taken into consideration. For the reason that the 3D ACA magnetic micro-robot and other 3D ACA microcapsules were located in a liquid environment and coded at the C10F18−H2O interface, the gravitation, buoyancy, and viscous fluid drag forces are taken into account for more precise trajectory planning and control. Therefore, the viscous fluid drag force can be given as Fvfd =

∫0

l

1 Cdρf A(x)v 2dx 2

(3)

where l represents the magnetic micro-robot radial length parallel to the direction of motion, Cd represents the drag coefficient that related to the magnetic micro-robot geometry, ρf represents the density of the liquid environment, A(x) represents the cross-sectional area of the magnetic micro-robot perpendicular to the direction of motion, and v represents the relative velocity of the magnetic micro-robot with respect to the fluid media.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.9b05517. TiOPc characteristics; formation of 3D heterogeneous ACA microcapsule; TiOPc-based light-induced electrodeposition setup; experimental setup of the home-made 4-DOFs electromagnetic coil system; SEM images of the 25671

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces



(10) Bian, W.; Liau, B.; Badie, N.; Bursac, N. Mesoscopic Hydrogel Molding to Control the 3D Geometry of Bioartificial Muscle Tissues. Nat. Protoc. 2009, 4, 1522−1534. (11) Ma, X.; Qu, X.; Zhu, W.; Li, Y. S.; Yuan, S.; Zhang, H.; Liu, J.; Wang, P.; Lai, C. S.; Zanella, F.; Feng, G. S.; Sheikh, F.; Chien, S.; Chen, S. Deterministically Patterned Biomimetic Human IpscDerived Hepatic Model via Rapid 3D Bioprinting. Proc. Natl. Acad. Sci. U.S.A. 2016, 113, 2206−2211. (12) Cui, J.; Wang, H.; Zheng, Z.; Shi, Q.; Sun, T.; Huang, Q.; Fukuda, T. Fabrication of Perfusable 3D Hepatic Lobule-Like Constructs Through Assembly of Multiple Cell Type Laden Hydrogel Microstructures. Biofabrication 2019, 11, No. 015016. (13) Liu, Y.; Wu, C.; Lu, H.; Yang, Y.; Li, W.; Shen, Y. Programmable Higher-Order Biofabrication of Self-Locking Microencapsulation. Biofabrication 2019, 11, No. 035019. (14) Nguyen, D. K.; Son, Y. M.; Lee, N. E. Hydrogel Encapsulation of Cells in Core-Shell Microcapsules for Cell Delivery. Adv. Healthcare Mater. 2015, 4, 1537−1544. (15) Xu, P.; Xie, R.; Liu, Y.; Luo, G.; Ding, M.; Liang, Q. Bioinspired Microfibers with Embedded Perfusable Helical Channels. Adv. Mater. 2017, 29, No. 1701664. (16) Zamanian, B.; Masaeli, M.; Nichol, J. W.; Khabiry, M.; Hancock, M. J.; Bae, H.; Khademhosseini, A. Interface-Directed SelfAssembly of Cell-Laden Microgels. Small 2010, 6, 937−944. (17) Du, Y.; Lo, E.; Ali, S.; Khademhosseini, A. Directed Assembly of Cell-Laden Microgels for Fabrication of 3D Tissue Constructs. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 9522−9527. (18) Yu, J.; Wang, Q.; Zhang, X. Effects of External Force Fields on Peptide Self-Assembly and Biomimetic Silica Synthesis. Appl. Surf. Sci. 2014, 311, 799−807. (19) Evers, C. H.; Luiken, J. A.; Bolhuis, P. G.; Kegel, W. K. SelfAssembly of Microcapsules via Colloidal Bond Hybridization and Anisotropy. Nature 2016, 534, 364−368. (20) Li, C. Y.; Wood, D. K.; Hsu, C. M.; Bhatia, S. N. DNATemplated Assembly of Droplet-Derived PEG Microtissues. Lab Chip 2011, 11, 2967−2975. (21) Li, M.; Yang, Q.; Liu, H.; Qiu, M.; Lu, T. J.; Xu, F. Capillary Origami Inspired Fabrication of Complex 3D Hydrogel Constructs. Small 2016, 12, 4492−4500. (22) Liljeström, V.; Seitsonen, J.; Kostiainen, M. A. Electrostatic Self-Assembly of Soft Matter Nanoparticle Cocrystals with Tunable Lattice Parameters. ACS Nano 2015, 9, 11278−11285. (23) Parker, R. M.; Zhang, J.; Zheng, Y.; Coulston, R. J.; Smith, C. A.; Salmon, A. R.; Yu, Z.; Scherman, O. A.; Abell, C. Electrostatically Directed Self-Assembly of Ultrathin Supramolecular Polymer Microcapsules. Adv. Funct. Mater. 2015, 25, 4091−4100. (24) Qi, H.; Ghodousi, M.; Du, Y.; Grun, C.; Bae, H.; Yin, P.; Khademhosseini, A. DNA-Directed Self-Assembly of Shape-Controlled Hydrogels. Nat. Commun. 2013, 4, No. 2418. (25) Chung, S. E.; Park, W.; Shin, S.; Lee, S. A.; Kwon, S. Guided and Fluidic Self-Assembly of Microstructures Using Railed Microfluidic Channels. Nat. Mater. 2008, 7, 581−587. (26) Chung, S. E.; Jung, Y.; Kwon, S. Three-Dimensional Fluidic Self-Assembly by Axis Translation of Two-Dimensionally Fabricated Microcomponents in Railed Microfluidics. Small 2011, 7, 796−803. (27) Liu, J.; Hwang, H. H.; Wang, P.; Whang, G.; Chen, S. Direct 3D-Printing of Cell-Laden Constructs in Microfluidic Architectures. Lab Chip 2016, 16, 1430−1438. (28) Wang, H.; Cui, J.; Zheng, Z.; Shi, Q.; Sun, T.; Liu, X.; Huang, Q.; Fukuda, T. Assembly of RGD-Modified Hydrogel Micromodules into Permeable Three-Dimensional Hollow Microtissues Mimicking in Vivo Tissue Structures. ACS Appl. Mater. Interfaces 2017, 9, 41669−41679. (29) Wang, H.; Li, J.; Cui, J.; Shi, Q.; Zheng, Z.; Sun, T.; Huang, Q.; Fukuda, T. Microrobotic Assembly of Shape-Customized ThreeDimensional Microtissues Based on Surface Tension Driven SelfAlignment. IEEE Trans. Nanotechnol. 2018, 17, 684−687. (30) Yue, T.; Nakajima, M.; Takeuchi, M.; Hu, C.; Huang, Q.; Fukuda, T. On-Chip Self-Assembly of Cell Embedded Micro-

core−shell ACA microcapsule; swelling behavior of ACA microcapsules with different membrane thicknesses; effect of the experimental parameters on the ACA microcapsule membrane; cell culture within the 3D ACA microcapsule (PDF) 3D square frame-shaped ACA microcapsule (MP4) Unevenly distributed Ca2+ ions in Ca-alginate hydrogel template results in local deformation of the ACA microcapsules (MP4) Swelling process of the ACA micro-architecture after liquefying (MP4)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (W.S.). *E-mail: [email protected] (Y.S.). ORCID

Yajing Shen: 0000-0001-5799-7524 Author Contributions

This work was supported by the National Science Foundation of China (U1813211), the Natural Science Foundation of Guangdong Province, China (Grant No. 2018A030313028), and the RGC General Research Fund of Hong Kong (CityU 11214817), and ShenZhen (China) Basic Research Project (JCYJ20160329150236426). Y.L. developed the system, fabricated the capsule, and conducted the experiments. G.L., Z.L., and W.S. conducted the cell culture related experiment and analysis. H.L. and Y.Y. developed the robot control system. Y.S. led the project. Notes

The authors declare no competing financial interest.



REFERENCES

(1) Lim, F.; Sun, A. M. Microencapsulated Islets as Bioartificial Endocrine Pancreas. Science 1980, 210, 908−910. (2) Ziv, K.; Gambhir, S. S. Bioengineering and Regenerative Medicine: Keeping Track. Nat. Mater. 2013, 12, 180−181. (3) Singh, A.; Peppas, N. A. Hydrogels and Scaffolds for Immunomodulation. Adv. Mater. 2014, 26, 6530−6541. (4) Ghaffarian, R.; Herrero, E. P.; Oh, H.; Raghavan, S. R.; Muro, S. Chitosan-Alginate Microcapsules Provide Gastric Protection and Intestinal Release of ICAM-1-Targeting Nanocarriers, Enabling GI Targeting in Vivo. Adv. Funct. Mater. 2016, 26, 3382−3393. (5) Naqvi, S. M.; Vedicherla, S.; Gansau, J.; McIntyre, T.; Doherty, M.; Buckley, C. T. Living Cell Factories - Electrosprayed Microcapsules and Microcarriers for Minimally Invasive Delivery. Adv. Mater. 2016, 28, 5662−5671. (6) Zhang, J.; Xu, X. D.; Liu, Y.; Liu, C. W.; Chen, X. H.; Li, C.; Zhuo, R. X.; Zhang, X. Z. Design of an “Active Defense” System as Drug Carriers for Cancer Therapy. Adv. Funct. Mater. 2012, 22, 1704−1710. (7) Zhang, L.; Chen, K.; Zhang, H.; Pang, B.; Choi, C. H.; Mao, A. S.; Liao, H.; Utech, S.; Mooney, D. J.; Wang, H.; Weitz, D. A. Microfluidic Templated Multicompartment Microgels for 3D Encapsulation and Pairing of Single Cells. Small 2018, 14, No. 1702955. (8) Chan, H. F.; Zhang, Y.; Leong, K. W. Efficient One-Step Production of Microencapsulated Hepatocyte Spheroids with Enhanced Functions. Small 2016, 12, 2720−2730. (9) Raof, N. A.; Padgen, M. R.; Gracias, A. R.; Bergkvist, M.; Xie, Y. One-Dimensional Self-Assembly of Mouse Embryonic Stem Cells Using an Array of Hydrogel Microstrands. Biomaterials 2011, 32, 4498−4505. 25672

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673

Research Article

ACS Applied Materials & Interfaces structures to Vascular-Like Microtubes. Lab Chip 2014, 14, 1151− 1161. (31) Du, Y.; Ghodousi, M.; Qi, H.; Haas, N.; Xiao, W.; Khademhosseini, A. Sequential Assembly of Cell-Laden Hydrogel Constructs to Engineer Vascular-Like Microchannels. Biotechnol. Bioeng. 2011, 108, 1693−1703. (32) Liu, Y.; Wang, Z.; Wang, X. AFM-Based Study of Fullerenol (C60(OH)24)-Induced Changes of Elasticity in Living SMCC-7721 Cells. J. Mech. Behav. Biomed. Mater. 2015, 45, 65−74. (33) Yang, S. M.; Tseng, S. Y.; Chen, H. P.; Hsu, L.; Liu, C. H. Cell Patterning via Diffraction-Induced Optoelectronic Dielectrophoresis Force on an Organic Photoconductive Chip. Lab Chip 2013, 13, 3893−3902. (34) Zhao, Y.; Lai, H. S.; Zhang, G.; Lee, G. B.; Li, W. J. Rapid Determination of Cell Mass and Density Using Digitally Controlled Electric Field in a Microfluidic Chip. Lab Chip 2014, 14, 4426−4434. (35) Williams, S. J.; Kumar, A.; Green, N. G.; Wereley, S. T. A Simple, Optically Induced Electrokinetic Method to Concentrate and Pattern Nanoparticles. Nanoscale 2009, 1, 133−137. (36) Yang, W.; Yu, H.; Li, G.; Wang, Y.; Liu, L. High-Throughput Fabrication and Modular Assembly of 3D Heterogeneous Microscale Tissues. Small 2017, 13, No. 1602769. (37) Xu, F.; Wu, C. A.; Rengarajan, V.; Finley, T. D.; Keles, H. O.; Sung, Y.; Li, B.; Gurkan, U. A.; Demirci, U. Three-Dimensional Magnetic Assembly of Microscale Hydrogels. Adv. Mater. 2011, 23, 4254−4260. (38) Tasoglu, S.; Diller, E.; Guven, S.; Sitti, M.; Demirci, U. Untethered Micro-Robotic Coding of Three-Dimensional Material Composition. Nat. Commun. 2014, 5, No. 3124. (39) Xu, F.; Inci, F.; Mullick, O.; Gurkan, U. A.; Sung, Y.; Kavaz, D.; Li, B.; Denkbas, E. B.; Demirci, U. Release of Magnetic Nanoparticles from Cell-Encapsulating Biodegradable Nanobiomaterials. ACS Nano 2012, 6, 6640−6649. (40) Jiang, D.; Park, S. Y. Light-Driven 3D Droplet Manipulation on Flexible Optoelectrowetting Devices Fabricated by a Simple SpinCoating Method. Lab Chip 2016, 16, 1831−1839. (41) Yang, S. M.; Yu, T. M.; Huang, H. P.; Ku, M. Y.; Hsu, L.; Liu, C. H. Dynamic Manipulation and Patterning of Microparticles and Cells by Using Tiopc-Based Optoelectronic Dielectrophoresis. Opt. Lett. 2010, 35, 1959−1961. (42) Dai, G.; Wan, W.; Zhao, Y.; Wang, Z.; Li, W.; Shi, P.; Shen, Y. Controllable 3D Alginate Hydrogel Patterning via Visible-Light Induced Electrodeposition. Biofabrication 2016, 8, No. 025004. (43) Liu, Y.; Wu, C.; Lai, H. S. S.; Liu, Y. T.; Li, W. J.; Shen, Y. J. Three-Dimensional Calcium Alginate Hydrogel Assembly via TiOPcBased Light-Induced Controllable Electrodeposition. Micromachines 2017, 8, 192. (44) Betz, J. F.; Cheng, Y.; Tsao, C. Y.; Zargar, A.; Wu, H. C.; Luo, X. L.; Payne, G. F.; Bentleyad, W. E.; Rubloff, G. W. Optically Clear Alginate Hydrogels for Spatially Controlled Cell Entrapment and Culture at Microfluidic Electrode Surfaces. Lab Chip 2013, 13, 1854− 1858. (45) Xie, Y. L.; Wang, M. J.; Yao, S. J. Preparation and Characterization of Biocompatible Microcapsules of Sodium Cellulose Sulfate/Chitosan by Means of Layer-by-Layer Self-Assembly. Langmuir 2009, 25, 8999−9005. (46) Ma, Y.; Zhang, Y.; Wang, Y.; Wang, Q.; Tan, M.; Liu, Y.; Chen, L.; Li, N.; Yu, W.; Ma, X. Study of the Effect of Membrane Thickness on Microcapsule Strength, Permeability, and Cell Proliferation. J. Biomed. Mater. Res., Part A 2013, 101, 1007−1015. (47) Zhang, W.; Zhao, S.; Rao, W.; Snyder, J.; Choi, J. K.; Wang, J.; Khan, I. A.; Saleh, N. B.; Mohler, P. J.; Yu, J.; Hund, T. J. A Novel Core-Shell Microcapsule for Encapsulation and 3D Culture of Embryonic Stem Cells. J. Mater. Chem. B 2013, 1, 1002−1009.

25673

DOI: 10.1021/acsami.9b05517 ACS Appl. Mater. Interfaces 2019, 11, 25664−25673