An Innovative Collagen-Based Cell-Printing Method for Obtaining

Mar 21, 2016 - ... Sungkyunkwan University (SKKU), Suwon 440-746, South Korea ... College of Medicine, Hallym University, Seoul 445-907, South Korea...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Biomac

An Innovative Collagen-Based Cell-Printing Method for Obtaining Human Adipose Stem Cell-Laden Structures Consisting of Core− Sheath Structures for Tissue Engineering MyungGu Yeo,† Ji-Seon Lee,‡ Wook Chun,‡ and Geun Hyung Kim*,† †

Department of Biomechatronic Engineering, College of Biotechnology and Bioengineering, Sungkyunkwan University (SKKU), Suwon 440-746, South Korea ‡ Department of Surgery, Hangang Sacred Heart Hospital, College of Medicine, Hallym University, Seoul 445-907, South Korea S Supporting Information *

ABSTRACT: Three-dimensional (3D) cell printing processes have been used widely in various tissue engineering applications due to the efficient embedding of living cells in appropriately designed micro- or macro-structures. However, there are several issues to overcome, such as the limited choice of bioinks and tailor-made fabricating strategies. Here, we suggest a new, innovative cell-printing process, supplemented with a core− sheath nozzle and an aerosol cross-linking method, to obtain multilayered cell-laden mesh structure and a newly considered collagen-based cell-laden bioink. To obtain a mechanically and biologically enhanced cell-laden structure, we used collagen-bioink in the core region, and also used pure alginate in the sheath region to protect the cells in the collagen during the printing and cross-linking process and support the 3D cell-laden mesh structure. To achieve the most appropriate conditions for fabricating cellembedded cylindrical core−sheath struts, various processing conditions, including weight fractions of the cross-linking agent and pneumatic pressure in the core region, were tested. The fabricated 3D MG63-laden mesh structure showed significantly higher cell viability (92 ± 3%) compared with that (83 ± 4%) of the control, obtained using a general alginate-based cell-printing process. To expand the feasibility to stem cellembedded structures, we fabricated a cell-laden mesh structure consisting of core (cell-laden collagen)/sheath (pure alginate) using human adipose stem cells (hASCs). Using the selected processing conditions, we could achieve a stable 3D hASC-laden mesh structure. The fabricated cell-laden 3D core−sheath structure exhibited outstanding cell viability (91%) compared to that (83%) of an alginate-based hASC-laden mesh structure (control), and more efficient hepatogenic differentiations (albumin: ∼ 1.7-fold, TDO-2: ∼ 7.6-fold) were observed versus the control. The selection of collagen-bioink and the new printing strategy could lead to an efficient way to achieve 3D cell-laden mesh structures that mimic the anatomical architecture of a patient’s defective region.

■

INTRODUCTION Recently, fabrication techniques for obtaining cell-laden structures have progressed rapidly due to various advances, which include homogeneous cell distribution and defined positioning of various cells in desired regions, as compared to scaffold-based cell-seeding methods.1,2 One of the various cellladen techniques, cell printing, based on an additive manufacturing method, has been used widely to achieve macroscopic cell-embedded structures using various biocompatible hydrogels and to enable efficient cell-positioning and successive deposition of cell-laden pieces in a multilayered manner, resulting in the generation of large-scale threedimensional (3D) tissue structures.3−9 For these reasons, the method may constitute an excellent cell-laden technique for tissue or organ regeneration. To successfully achieve a cell-printing process, one of the major factors has been the hydrogel matrix (bioink) containing the cells. An appropriate hydrogel not only provides efficient process-ability to construct the 3D macroporous cell-laden © XXXX American Chemical Society

structure but also performs functions as an excellent biological component for the laden cells. Thus, the cell-containing hydrogel requires various properties, including cytocompatibility, controllable biodegrabability, efficient printability, and mild cross-link-ability under “safe” conditions to construct the macroscopic 3D structure.5 The most representative hydrogels used in cell-printing processes are based on calcium alginate because of its easy process-ability, rapid gelation, and nontoxic properties.10−13 However, because most recommended hydrogels are not major components of natural extracellular matrixes (ECMs), and even those that are, are not adequate to build a microenvironment that involves cell−cell interactions, better cell-printing processes using better ECM-based hydrogels are required.14 Received: December 30, 2015 Revised: March 17, 2016

A

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

of the collagen-based cell-laden structures, we measured various rheological properties for various cross-linking conditions and assessed in vitro biological activities: cell viability, DAPI/ phalloidin analysis, and cell proliferation. Additionally, the ability of the hASCs laden in the cell structure to undergo hepatogenic differentiation was assessed with the expression of liver-specific genes. From the results, we can see that the new processing technique using collagen-based bioink may have great potential for creating 3D cell-laden macroporous structures that mimic the properties of native tissue.

To overcome these challenges, new cell-printing strategies using new ECM-derived hydrogels have been suggested to achieve macro-scale 3D tissue structures.15−17 Recently, Bertassoni et al. suggested a new strategy for cell-printing with cell-laden methacrylated gelatin (GelMA) hydrogels.17 They processed fibroblast-embedded GelMA with various concentrations (7 to 15%) and cell-densities (1 × 106 to 6 × 106 cells mL−1) and showed that successful cell-printing was closely related to the elastic modulus of the hydrogel. Additionally, we recently developed a new approach for achieving cell-printed 3D porous cell-laden structures using ECM-based bioink consisting of collagen/ECM and alginate.16 As a result, cellular activities, including hepatogenic differentiation of human adipose stem cells (hASCs), in the cell blocks were improved significantly, compared with a pure alginate-based cell-laden structure. Generally, collagen-based scaffolds have been revealed capability to promote cell and tissue attachment and growth.18−22 For these unique properties, collagen is widely used as a biomedical scaffold for regenerating bone, skin, and cartilage.23 However, despite various advantageous properties of collagen, limitations such as unsuitable mechanical properties, low processability in terms of control of the micropore structure, and high cost have prevented collagen from usage in various applications. In a previous study, we proposed a 3D collagen-based structure fabricated by a 3D dispensing system supplemented with a low temperature system.24,25 However, the dispensed collagen scaffolds should be cross-linked using toxic chemical reagents like EDC or glutaraldehyde. The crosslinking process using toxic chemicals has been one of the big obstacles of the usage of collagen as a cell-laden bioink. Recently, Suri et al. fabricated collagen-based cell-laden 3D structure using a photo-cross-linking process.26 However, the cell-laden structure was fabricated by injecting the solution (collagen and cells) into a predesigned silicone mold, so that the structure has some shortcomings, such as nonhomogeneous cell distribution and low porous structure. In addition, several researches on fabricating cell-laden core−sheath structure using microfluidic devices have been investigated.14,27,28 Lee et al. developed the microfluidic chip process, and fabricated HIVE78 cell-laden two-dimensional (2D) wound alginate hollow fiber matrix.28 They were able to fabricate cell-laden microfibers mimicking microvascularized structure, but the 3D shapeability can be an obstacle of the application. Also, previously, by using a core−sheath nozzle and low temperature printing process, we designed collagen-based scaffolds, but cells were not simultaneously laden due to the harsh cross-linking and fabricating process.9,29−31 In this study, we suggest another strategy for a cell-printing process to obtain highly porous collagen-based 3D cell constructs by using a combination method of a 3D cellprinting system and an aerosol tentative cross-linking process. Here, a homemade cell printer with a core/sheath nozzle was used to dispense microsized core/sheath struts consisting of a core (cell-laden collagen) and sheath (pure alginate) region. The cell-printing method overcomes the restrictions related to the limited collagen-based hydrogels that have low processing ability (due to slow cross-linking ability and chemical crosslinking agents toxic to the embedded cells) to fabricate a 3D macroporous structure. Using this technique, osteoblast-like (MG63) cells and human adipose stem cell (hASC)-laden collagen were used to dispense with collagen to achieve 3D mesh structures. To assess the printability and cellular activities

■

EXPERIMENTAL SECTION

Materials and Bioinks. Osteoblast-like cells (MG63; ATCC, Manassas, VA, USA) and human adipose-derived stem cells (hASCs; Anterogen Corp., South Korea) were used in cell-printing processes. Low-viscosity, high-G-content LF10/60 alginate (FMC BioPolymer, Drammen, Norway), and type I collagen from porcine tendon (Matrixen-PSP; Bioland, Cheonan City, South Korea) were used in the cell-printing processes. The mixture of cells and collagen (2 wt %), mixed using a three-way stopcock tool at a density of 5 × 106 cells mL−1 for MG63 and 6.7 × 106 cells mL−1 for hASCs, was used as a collagen-based bioink in the core region of the core−sheath nozzle. In addition, a cells-laden alginate (4 wt % in phosphate buffered saline (PBS)) having the same cell density was used as a control bioink. Before mixing the cells, the alginate solution was slightly cross-linked using 0.5 wt % CaCl2 solution at a 7:3 ratio. The mixture of cells and alginate solution was then loaded into a syringe to be printed. Fabrication of 3D Porous Cell-Laden Mesh Structures Using a 3D Cell-Printer. A computer-controlled three-axis printing system (DTR2-2210T, Dongbu Robot, Bucheon, South Korea) was used with a dispenser, connected with a single nozzle and core−sheath nozzle, to obtain multilayered cell-laden mesh structures. A humidifier (Tess7400; Paju, South Korea) was used for the aerosol cross-linking process. The pneumatic pressures in the core and sheath nozzle of the cell-printing process were controlled separately with two different dispensers. Each layer of dispensed bioinks was dispensed on top of the existing one, in the planar direction, perpendicular to the lower layer, forming a porous mesh structure. Storage Modulus Change for the Concentration of CaCl2 Solution of an Aerosol Process. The rheological properties (i.e., storage modulus, G′) of the alginate solution for time sweep were evaluated using a rotational rheometer (Bohlin Gemini HR Nano, Malvern Instruments, Surrey, UK) equipped with cone-and-plate geometry (40 mm in diameter; with a cone angle of 4° and a gap of 150 μm). A dynamic time sweep was carried out with 1% strain and at 32 °C within the linear viscoelastic region to compare the modulus change of alginate solution with no aerosol cross-linking process and an aerosol process with various concentrations of calcium chloride solution. Characterization of the 3D Structure Fabricated Using the Cell-Laden Structures. The pore structure of the 3D cell-laden structures was observed using an optical microscope (Model BX FM32; Olympus, Tokyo, Japan) and a scanning electron microscope (SEM, SNE-3000M, SEC, Inc., South Korea). The pore size of the mesh structures was defined as the separation between the parallel cylindrical struts. The compressive stress−strain curves were characterized using a universal testing machine (Top-tech 2000; Chemilab, Seoul, South Korea) in compressive mode. The cell-laden mesh structures were cut into a small cylindrical shape with a biopsy punch (diameter = 6 mm) and compressed in the thickness direction. Stress−strain curves were recorded at a compression rate of 0.5 mm s−1 with samples with 6 mm of diameter and 2.2 mm thick. Elastic modulus was measured from the slope fit to the stress−strain curve at 15% strain. All data are expressed as the mean ± standard deviation, and the measurements were repeated five times. A Fourier-transform infrared (FT-IR) spectrometer (model 6700; Nicolet, West Point, PA, USA) was used to observe the chemical structures of the released components from the cell-laden mesh B

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 1. Schematic showing the cell-printing process with a core−sheath nozzle using a collagen-based bioink (core) and pure alginate (sheath). (a) The cell-printing process was supplemented with a core−sheath nozzle and an aerosol cross-linking process using 10 wt % calcium chloride solution (core: cell-laden collagen, sheath: pure alginate). (b) Schematic images of the finally fabricated product with the modified cell-printing process (AC structure) and control, fabricated with a general cell-printing process using a cell-laden alginate solution. structure. IR spectra represent the mean of 30 scans at 500−4000 cm−1 at a resolution of 8 cm−1. In Vitro Cell Culture and MTT Assay. The mesh structures were dispensed with bioinks containing MG63s or hASCs, and they were cultured and maintained in α-modified minimum essential medium (αMEM) containing 10% fetal bovine serum (FBS) and 1% antibiotic/ antimycotic (Cellgro, Mediatech, Manassas, VA). The structures were incubated in an atmosphere of 5% CO2 at 37 °C, and the medium was changed every second day. The proliferation of viable cells was determined using the MTT cell proliferation assay (Cell Proliferation Kit I; Boehringer Mannheim). The assay is based on the cleavage of the yellow tetrazolium salt, MTT, via mitochondrial dehydrogenases in viable cells, to produce purple formazan crystals. Cell-laden structures were incubated in a 0.5-mg mL−1 MTT solution for 4 h at 37 °C. The volume of the cultured sample for measurement was 6 × 6 × 2.4 mm3 and the number of the stacked layers was two. The amount of the MTT solution during the incubation was 400 μL. Only 100 μL of the supernatant around the cell-laden scaffold in the MTT solution was selected as a reading sample and moved to a 96-well plate. Cell-laden structure samples were not recycled, and different samples were prepared for different time points. The absorbance at 570 nm was measured using a microplate reader (EL800; Bio-Tek Instruments). Four samples were tested for each incubation period, and each test was performed in triplicate. Live/Dead Cell Assay and DAPI/Phalloidin Analysis. To measure the cell viability of the dispensed single struts and mesh structures, they were exposed to 0.15 mM calcein-AM, and 2 mM ethidium homodimer-1 for 45 min in an incubator. The stained structure was then analyzed using a microscope (TE2000-S; Nikon, Tokyo, Japan) equipped with an epifluorescence attachment and a SPOT RT digital camera (SPOT Imaging Solutions, Sterling Heights, MI, USA). To evaluate the initial cell viability after 4 h and 1 day, the numbers of live and dead cells were counted using the ImageJ software (NIH, Bethesda, MD, USA). The ratio of the number of live cells to the number of total cells (i.e., live and dead cells) was calculated, and this ratio was normalized to the cell viability determined using trypan blue (Mediatech, Herndon, VA, USA) before printing with the bioinks. Additionally, the 3D mesh structures were analyzed with diamidino2-phenylindole (DAPI; 1:100, Invitrogen, Carlsbad, CA, USA) staining. Fluorescence staining was performed to characterize the cell nuclei in the samples (Invitrogen). Alexa Fluor 568 phalloidin (1:100, Invitrogen) was used to visualize the actin cytoskeleton of the cells in the structure by fluorescence microscopy. In Vitro Differentiation of hASCs to Hepatocytes and Staining of Released Cells. hASCs were provided by Anterogen Corp. and were cultured in low-glucose Dulbecco’s modified Eagle

Medium (DMEM; cat.: SH30021.01, Hyclone) containing 0.1% gentamicin and 10% FBS. hASCs laden in a 3D structure were incubated in basal medium (low-glucose DMEM containing 0.1% gentamicin, 10% FBS) after the construction of 3D cell-blocks. The next day, the 3D cell-blocks were washed with Hank’s balanced salt solution (HBSS) twice and changed to low-glucose DMEM containing 0.1% gentamicin, 2% FBS, 20 ng/mL EGF (CYT-217, ProSpec), and 10 ng/mL basic fibroblast growth factor (bFGF) (CYT-218, ProSpec) for 2 days. Then, the medium was changed to low-glucose DMEM containing 0.1% gentamicin, 2% FBS, 1 × ITS (insulin−transferrin− selenium supplement), 10 ng/mL Oncostatin-M (OSM) (CYT-231, ProSpec), 10 ng/mL bFGF, 20 ng/mL HGF(CYT-244, ProSpec), 1 μM dexamethasone, 5 mM nicotinamide, and 0.1% dimethyl sulfoxide (DMSO). The medium was replaced every 2−3 days. Hepatogenic differentiation was assessed by reverse transcription polymerase chain reaction (RT-PCR) analysis of liver-specific genes. After 20 days of cell culture, for immunofluorescence staining, released cells were placed on a coverslip. Cells were washed twice with PBS, fixed in 4% paraformaldehyde (PFA), washed with PBS, and then incubated for 3 min with 0.1% (v/v) Triton X-100 in PBS. Cells were then incubated with blocking solution (1:10, ab126587, Abcam) for 30 min and 0.05% TBS-T containing antibodies against albumin (1:100, ab135575, Abcam) for 90 min at room temperature (RT). After washing with 0.05% TBS-T in PBS, cy2-conjugated fluorescent secondary antibodies (Jackson ImmunoResearch) and DAPI were added at dilutions of 1:200. The cells were again washed with 0.05% TBS-T three times and imaged under a fluorescent microscope (Olympus IX81). RNA Isolation and Quantitative Real-Time PCR Analysis for Measuring Hepatogenic Genes. Total RNA from the cultured 3D cell-blocks was isolated using the Trizol reagent (Life Technologies, Inc., Carlsbad, CA) according to the manufacturer’s instructions. Gene-specific primers were as follows: human albumin (ALB) (forward: 5′-GTCACCAAATGCTGCACAGA-3′, reverse: 5′ACGAGCTCAACAAGTGCAGT-3′), human tryptophan 2,3-dioxygenase (TDO2) (forward: 5′-GTGTGCATGGTGCACAGAAT-3′, reverse: 5′-GGGTTCATCTTCGGTATCCA-3′) and housekeeping gene human β-actin (forward: 5′-GTCCTCTCCCAAGTCCACAC3′, reverse: 5′-GGGAGACCAAAAGCCTTCAT-3′). All amplifications were conducted in a premixture (20 μL) containing 500 nmol/L of gene-specific primers, 2× SYBR, and 6 μL of template, under the following conditions: denaturation at 95 °C for 5 min, followed by 40 cycles of 95 °C for 10 s, 58 °C for 15 s, and 72 °C for 15 s, with a final extension at 72 °C for 5 min. The reactions were carried out in a Roche LC480. C

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 2. Effects of an aerosol cross-linking process during cell printing on designing the core−sheath structure. Time periodic optical images describing the diffusion of cell-laden collagen (mixed with rhodamine (red color)) in the core region (a) without and (b) with using the aerosol cross-linking process (10 wt % of calcium chloride and flow rate = 0.93 ± 0.12 mL min−1). (c) Comparison of diffusing range of rhodamine in both processes.

Figure 3. Effect of weight fractions of calcium chloride solution, used in an aerosol cross-linking process, on shapeability in fabricating a mesh structure. (a) Optical images of porous mesh structure for various weight fractions of the cross-linking agent under a constant flow rate (0.93 ± 0.12 mL min−1). Comparison of (b) strut diameter and (c) pore size of the fabricated mesh structures. (d) The storage modulus, G′, of alginate before and after using the aerosol process for various weight fractions of the cross-linking agent. (e) Comparison of the moduli at 7 min. Statistical Analysis. Data presented are expressed as means ± SD. Single-factor analysis of variance (ANOVA) was used as the statistical test. The significance level was set at p < 0.05 (*).

consisted of a mixture of cells and collagen (2 wt %) to induce high cellular activity. During the process, simultaneously, an aerosol process using calcium chloride solution was applied to tentatively cross-link the alginate component in the sheath region, so that the cell-laden collagen component could be stably retained in the core region of the cylindrical core−sheath struts, as shown in Figure 1b. In this work, we fabricated two different types of cell-laden porous structure: AC structure and control. As shown in the schematic images of Figure 1b, the AC

■

RESULTS AND DISCUSSION In Figure 1a, a schematic describing the cell-printing process using aerosol-assisted 3D printing and the core−sheath nozzle is shown. In the image, the sheath region was dispensed with a pure alginate (4 wt %) to protect the cell-laden bioink, which D

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 4. Fabricated mesh structures consisting of core−sheath structure and FT-IR analysis. (a) Fabricated mesh structure without and with the aerosol process. (b) Optical images showing the released components from the cell-laden mesh structure after dipping for 1 day and (c) FT-IR results of the released components.

cell-laden structure consisted of a core (cell-laden collagen bioink) and sheath (pure alginate) component, while the control structure consisted of cell-laden alginate bioink, which was fabricated conventionally using general cell printing with a single nozzle. Selection of Optimal Processing Conditions to Obtain a Core−Sheath Structure. To observe the effect of the aerosol process using a calcium chloride solution (10 wt % CaCl2 and flow rate = 0.93 ± 0.12 mL min−1) on the crosslinking ability of the alginate component in the sheath region, we compared it with a process not using an aerosol process. Figure 2a,b shows the droplet obtained using the core−sheath nozzle under pneumatic pressures (core = 12 kPa and sheath = 240 kPa) during 1 s. In the images, the core region was occupied with a mixture of cell-laden collagen and rhodamine, while in the sheath region was filled with pure alginate. As shown in the result in Figure 2a, the red color (rhodamine) diffused rapidly from the core region to the sheath region. However, on exposing the droplet of the aerosol to calcium chloride solution for 20 s, the diffusion was reduced significantly due to the surface cross-linking of the alginate region of the droplet. The time-periodic images were analyzed using the percent of diffusion of rhodamine from the center of the droplet for the processes with and without an aerosol solution (Figure 2c). As expected, in the aerosol cross-linking process, the diffusion of rhodamine in the core region was

saturated at 15 s in the cross-linking conditions using the aerosol solution. To analyze more precisely the effects of the concentration of the calcium chloride solution on the aerosol process, we applied various weight fractions of CaCl2 solution to fabricate the cylindrical struts consisting of core (cell-laden collagen) and sheath (pure alginate). Figure 3a shows the optical images for the fabricated mesh structures subjected to various weight fractions (5, 7.5, 10, and 12.5 wt %) of the aerosol cross-linking solution. In this test, we fixed the flow rate of the aerosol solution and the nozzle moving speed (2 mm s−1). As shown in the images, for 5 and 7.5 wt %, the border of the core and sheath region was not visible, due to the low degree of crosslinking in the alginate region in the sheath, while with increasing the weight fraction of the aerosol solution above 10 wt %, the border was clearly seen, due to sufficient crosslinking of the alginate. As shown in the optical images, the diameter of the cell-laden cylindrical struts was enlarged, compared with the initial designed diameter of the struts. This was also because of the inadequate degree of cross-linking of the alginate component in the sheath region. A detailed analysis of the structural geometry for the weight fraction of the aerosol is shown in Figure 3b,c. As expected, above 10 wt % of the aerosol solution, a stable shape ability (strut diameter and pore size) of the cell-laden mesh structure was achieved. Generally, the cross-linking ability of the solution can be evaluated indirectly with the rheological data (e.g., the storage E

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 5. Optical images showing the core and sheath region of the bioink extruded from the nozzle tip for various core/sheath pneumatic pressures and compressive properties of the fabricated mesh structures having different core regions (collagen bioink). (a) Optical images of the bioink extruded from the nozzle for various pneumatic pressures in the core (6, 12, 24, and 36 kPa) and (b) measured size of the core and sheath region. (c) Optical images of the bioink for various sheath pressures (50, 100, 150, and 200 kPa) and (d) measured size of the core and sheath region. (e) Optical and (f) SEM images of the mesh structure fabricated with the core (12 kPa) and sheath (240 kPa) pressure. (g) Compressive stress−strain curves for various mesh structures and (h) elastic modulus.

modulus, G′). To examine the cross-linking ability of the alginate for the various weight fractions of the aerosol solution, we measured the storage modulus. First, we measured the pure alginate not exposed to the aerosol solution for 3 min and then we applied the aerosol to the alginate solution (flow rate =0.93 ± 0.12 mL min−1 and 15 s). After the exposure of the aerosol conditions, we measured the storage modulus. Figure 3d shows the storage modulus for the alginate solution before and after the aerosol-cross-linking process. With increasing weight fraction of the CaCl2 solution, the modulus was enhanced significantly, and the increasing value reached the saturation point at 10 wt % CaCl2 solution (Figure 3e). Figure 4a shows the fabricated cell-laden core−sheath structures with and without the aerosol cross-linking process. As shown in the optical images, with increasing weight fraction of the cross-linking agent, a stable mesh structure was obtained. In addition, we simply measured the shielding ability of the cross-linked alginate-sheath region to encapsulate the cell-laden collagen (core region) using dipping in PBS solution. Figure 4b shows the optical images of the cell-laden structures after dipping for 1 day. As shown in the images, in the structures cross-linked with 5 and 7.5 wt % calcium chloride solutions,

released material from the structures was observed (inset image, Figure 4b). The released material was evaluated with FT-IR analysis (Figure 4c). In the spectrum, the pure collagen showed several peaks corresponding to amide I band (1635 cm−1), amide II bands (1543 and 1452 cm−1), and amide III band (1239 cm−1) and also for pure cross-linked alginate, the spectra indicated carboxyl peaks (1602 cm−1, symmetric COO− stretching vibration) and 1412 cm−1 (asymmetric COO− stretching vibration). Based on the spectra, we can estimate that the released material from the structures was a mixture of collagen and alginate due to the nonstably cross-linked alginate sheath region. From these results, we can conclude that the optimal concentration of the cross-linking agent for the selected aerosol cross-linking process was likely to be 10 wt %. However, this value may be limited to this selected aerosol process. Generally, the volume flow rate in the core region can straightforwardly manipulate the occupying region of the core and sheath structure. In particular, the controllability of the core region in the core−sheath structure is an important parameter because the core area not only provides the amount of embedded cells but also directly influences the mechanical F

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 6. Optical and fluorescence images and cellular activities of the cell-laden mesh structures. (a) Optical and fluorescence (live = green, dead = red) images of (a) MG63-laden structure (AC structure) having core (cell-laden collagen)-sheath (alginate) structure and (b) a control with hASCsladen alginate. Live−dead images for the mesh structures ((c) AC structure and (d) control) after cell culturing for 1 and 7 days. (e) Initial cell viability of the cell-laden structures at 1 and 14 days. (f) MTT assay for the cell-laden structures.

properties of the finally fabricated 3D cell-laden structure. In this work, we studied the processing conditions for obtaining a cell-laden structure with a variety of core diameters by handling the pneumatic pressure (P−C) forced in the core region. In the test, we fixed the processing temperature at 32 °C, because the environmental temperature can directly influence the shear viscosity of the alginate and collagen bioink. Also, the pressure in the sheath region was varied slightly to achieve a similar strut whole diameter, and the same aerosol process condition of the CaCl2 solution (10 wt % and 0.93 ± 0.12 mL min−1) was applied during the process. Figure 5a shows the optical images in the vicinity of the nozzle for various P−Cs (6, 12, 24, and 36 kPa) and slightly different sheath pressures (250, 240, 230, and 220 kPa, respectively). In the images, with increasing pressure in the core region, the region of red color (rhodamine), indicating the cell-laden collagen, increased stably because the interfacial tension between the core and the sheath solutions may be low. However, at 36 kPa in the core region, unstable mixing of the core and the sheath solutions occurred because the viscous drag supplied by the alginate solution was not enough to sustain the collagen bioink within the core region. This means that the alginate cannot fully encapsulate the rapidly moving collagen solution consistently, making the core−sheath process unstable.32 Figure 5b shows the various diameters (134 ± 45 μm, 284 ± 43 μm, and 447 ± 41 μm) of the core region for various P−Cs, but similar diameters (968 ± 76 μm) of the whole cylindrical struts. Using this process, we can obtain three cell-laden mesh structures fabricated at three different core pressures (6, 12, and 24 kPa). In addition, we measured the effect of various sheath pressures (50, 100, 150, and 200 kPa) under a constant core pressure (12 kPa) on the core and sheath size. Figure 5c,d shows the optical images and measured sizes.

As shown in the images, as the sheath pressure increased, the diameter of both core and sheath region was gradually increased due to the interfacial drag of the higher volume flow rate of the alginate solution in the sheath region. However, for the low sheath pressure, the core solution occupied a large portion of the core−sheath strut, eventually causing a clog. Figure 5e shows an optical image of the cell-laden mesh structure fabricated using a core pressure of 12 kPa, and the cell-laden collagen region resided well within the alginate sheath. The fabricated mesh structure was freeze-dried, and Figure 5f shows the surface and cross-sectional SEM images. To observe the mechanical properties of the mesh structures having different core regions, we conducted compressive stress−strain tests (Figure 5g). Using the stress−strain curves, the elastic modulus was measured for the structures (Figure 5h). As shown, with increasing concentrations of collagen, the mechanical stability decreased significantly, due to the enlarged non-cross-linked collagen region of the cell-laden struts. However, the modulus of the cell-laden structures using pneumatic pressures of 6 and 12 kPa was not significant. Based on these mechanical properties and encapsulating ability, we selected the core and the sheath pressures as 12 and 240 kPa, respectively. Fabrication and Cellular Activities of Cell-Laden Mesh Structure Consisting of Microsized Core−Sheath Struts. Generally, the initial cell viability of the cell-printing process is a significant parameter because the viability can directly influence various cellular activities including cell proliferation and even differentiation for culturing periods.33 To observe the initial cell viability of the core−sheath cell-laden structure, we used a cellladen alginate structure, which was fabricated using a cellprinting process, and the same aerosol cross-linking method as a control. The control consisted only of pure alginate and cells, G

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 7. Optical and fluorescence images of the hASCs-laden mesh structure and control structure. (a) Schematic images for designing AC structure and control. Optical and fluorescence (live = green, dead = red) images of (b) hASCs-laden structure (AC structure) having core (cellladen collagen)-sheath (alginate) structure and (c) a control with hASCs-laden alginate. (d) Initial cell viability at 4 h and 1 day, (e) MTT assay, and (f) proliferation rate of viable cells for the AC structure and control laden with hASCs.

were safely printed, and proliferated (Figure 6d). However, as the viabilities were evaluated quantitatively, the cell viabilities at 1 and 14 days between the AC structure and control were significantly different (p < 0.05) because the cells in the control structure can directly contact high wall shear stress within the nozzle and the calcium chloride solution during the aerosol process (Figure 6e). To observe the metabolic activity of the cell-laden structures for longer culture periods, we performed the MTT assay at 1, 3, and 7 days (Figure 6f). The AC cellladen structure showed outstanding metabolic activity compared with the control structure. This was because the cells residing in the core region were more safely extruded, and the collagen, as the bioink, operated as an outstanding cellactivating material to induce efficient microenvironmental cellto-cell interactions.35 Application of the Core−Sheath Method for Obtaining Human ASCs-Laden 3D Porous Structures. Recently, hASCs have been used widely in various tissue regenerations and cell therapy because of their multipotent and efficient stimulation into various cell types, such as chondrogenic, osteogenic, and even mesenchymal lineages, and the hASCs can also stimulate angiogenesis and reduce inflammation.36,37 Recently, there has been interest in the possibility of hASCs undergoing hepatogenic differentiation. According to Wang et al.38 hASCs seeded in a porous poly lactide-co-glycolide (PLGA) scaffold were implanted in a rat model, and the cells loaded in the matrix under a hepatic-inducing media

without using a core−sheath nozzle. To identify the live and dead cells, the printed mesh structure was stained using calceinAM and ethidium homodimer-1, respectively. In the fabricated mesh structure, the MG63 cells had a density of 5.0 × 106 cells mL−1, and the concentrations of alginate in the sheath and cellladen collagen in the core region were fixed at 4 and 2 wt %, respectively. We applied the same aerosol cross-linking process. Figure 6a,b shows optical and live/dead images for the cellladen mesh structure (AC structure, 15 × 15 × 2.4 mm3) constituted of cylindrical struts (the sheath (alginate) and core (collagen+cells) region), and the control structure (pure cellladen alginate). According to Haraguchi et al., to efficiently transport nutrients and waste, the maximum thickness of a cellencapsulated block should be below 100 μm, although the value can be varied for the degree of cross-linking and concentration of cell capsulation material.34 As shown in the surface and cross-sectional fluorescence images of Figure 6a, the cell-laden collagen was well encapsulated within the alginate sheath and the thickness of the alginate was about 106 ± 36 μm. Although the thickness of alginate sheath may be slightly higher than the recommended 100 μm, we think the thickness may be sufficient to sustain the metabolic activities of the cells laden in the collagen region for the long culture period. In addition, for the cell viability after culturing 1 and 7 days, the cells in the core region were completely safe, due to sufficient protection of the alginate during the printing process, and they proliferated well (Figure 6c). Also, for the control, the cells in the alginate region H

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 8. Fluorescence, SEM, optical, and DAPI/phalloidin images of the hASCs-laden mesh structures and hepatogenic differentiation. Fluorescence (live/dead) and cross-sectional SEM image of (a) AC structure and (b) control after cell-culturing 7 days. Optical and DAPI/ phalloidin images of (c) AC structure and (d) control after cell-culturing 20 days. Expression levels of hepatocyte-specific genes (e) albumin and (f) TDO-2 using real-time RT-PCR on day 20.

cell proliferation rate of AC-structure (C0 = 0.095, k = 0.068) exhibited less than that of control (C0= 0.089, k = 0.058) (Figure 7f). As expected, the cell viability and cell proliferation of the AC structure were significantly higher than those of the control. These results were similar to those of the MG63-laden structures of Figure 6e,f. Figure 8a,b shows live/dead images after cell-culture 7 days and Figure 8c,d shows optical and fluorescence images of released cells after 20 days of culture from hASCs-laden structures (AC structure and control). As shown in the live/ dead images, the cells in the AC-structure and control were well alive, and the structure consisting of the struts was well sustained during the cell-culture period. Release of cells was much greater from structures fabricated using the core (cellladen collagen)/sheath (pure alginate) nozzle system versus the control. Figure 8b shows DAPI/phalloidin images after 20 days and the analysis of the nuclei number and F-actin area, respectively, and they confirm that hASCs were more efficiently distributed on the strut surface in the AC structure than in the control. These results showed that the metabolic function of the cells in the AC structure was much more active during the culture period than that in the control structure.

proliferated well and efficiently differentiated into hepatocytelike phenotypes. To expand the applicability of the core−sheath process, we fabricated hASCs-laden 3D porous structures in which hASCs-laden collagen resided in the core region and pure alginate was laden in the sheath region of the cylindrical struts. Also, we fabricated 3D hASCs-laden alginate structures as a control. To obtain a cell-laden mesh structure, we applied the same processing conditions (aerosol and printing process) that were used in fabricating the MG63-laden mesh structure. The density of hASCs in the cell-printing process was 6.7 × 106 mL−1. Figure 7a shows a schematic image of the cell-laden mesh structure consisting of two porous layers (strut diameter = 400 μm and pore size = 500 μm) with and without a core-region. Figure 7b,c show the optical and fluorescence images describing live−dead cells for the AC structure and control using hASCs, respectively. As shown in the AC structure laden with hASCs, only the cells in the core region looked healthy. In Figure 7d−f, the cell viability, MTT assay, and proliferation of viable cells were compared. The cell proliferation rate (k) was calculated using the MTT data (i.e., the optical density, C(t), as a function of culture time (t)) from the equation, C(t) = C0ekt, where C0 is the initial optical density and k is the proliferation rate.39 The I

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Education, Science, and Technology (MEST) (Grant No. NRF-2015R1A2A1A15055305) and also a grant from the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea (Grant No. HI15C3000).

To observe differentiation in the hASCs-laden structures, ALB and TDO2 expression were measured after 20 days of culturing in the hepatogenic medium (Figure 8c,d). In the results, “NT” indicates undifferentiated hASCs, which were not treated with hepatogenic medium. The hASCs from the fabricated structures were found to express liver-specific genes. However, the level of liver-specific gene expression in the AC structure was much higher than those in the control. This indicates that the modified cell-printing system provided more favorable hepatogenic differentiation conditions for hASCs than the control, because the process can use biocompatible collagen-bioink and protective alginate layer around the cell-laden bioink to improve cell viability during the printing process.

■

(1) Kapur, S. K.; Wang, X.; Shang, H.; Yun, S.; Li, X.; Feng, G.; Khurgel, M.; Katz, A. J. Biofabrication 2012, 4, 025004. (2) Yao, R.; Zhang, R.; Luan, J.; Lin, F. Biofabrication 2012, 4, 025007. (3) Fedorovich, N. E.; Schuurman, W.; Wijnberg, H. M.; Prins, H. J.; van Weeren, P. R.; Malda, J.; Alblas, J.; Dhert, W. J. A. Tissue Eng., Part C 2012, 18, 33−44. (4) Fedorovich, N. E.; Alblas, J.; Hennink, W. E.; Oner, F. C.; Dhert, W. J. A. Trends Biotechnol. 2011, 29, 601−606. (5) Murphy, S. V.; Atala, A. Nat. Biotechnol. 2014, 32, 773−85. (6) Cohen, D. L.; Malone, E.; Lipson, H.; Bonassar, L. J. Tissue Eng. 2006, 12, 1325−1235. (7) Ahn, S. H.; Lee, H.; Bonassar, L. J.; Kim, G. H. Biomacromolecules 2012, 13, 2997−3003. (8) Ouyang, L.; Yao, R.; Mao, S.; Chen, X.; Na, J.; Sun, W. Biofabrication 2015, 7, 044101. (9) Ahn, S. H.; Lee, H.; Lee, J.-S.; Yoon, H.; Chun, W.; Kim, G. H. Sci. Rep. 2015, 5, 13427. (10) Guillotin, B.; Souquet, A.; Catros, S.; Duocastella, M.; Pippenger, B.; Bellance, S.; Bareille, R.; Rémy, M.; Bordenave, L.; Amédée, J.; Guillemot, F. Biomaterials 2010, 31, 7250−7256. (11) Duan, B.; Hockaday, L. A.; Kang, K. H.; Butcher, J. T. J. Biomed. Mater. Res., Part A 2013, 101, 1255−1264. (12) Jia, J.; Richards, D. J.; Pollard, S.; Tan, Y.; Rodriguez, J.; Visconti, R. P.; Trusk, T. C.; Yost, M. J.; Yao, H.; Markwald, R. R.; Mei, Y. Acta Biomater. 2014, 10, 4323−4331. (13) Ahn, S. H.; Lee, H.; Lee, E. J.; Kim, G. H. J. Mater. Chem. B 2014, 2, 2773−2782. (14) Onoe, H.; Okitsu, T.; Itou, A.; Kato-Negishi, M.; Gojo, R.; Kiriya, D.; Sato, K.; Miura, S.; Iwanaga, S.; Kuribayashi-Shigetomi, K.; Matsunaga, Y. T.; Shimoyama, Y.; Takeuchi, S. Nat. Mater. 2013, 12, 584−590. (15) Pati, F.; Jang, J.; Ha, D.-H.; Kim, S. W.; Rhie, J.-W.; Shim, J.-H.; Kim, D.-H.; Cho, D.-W. Nat. Commun. 2014, 5, 3935−3945. (16) Lee, H. J.; Kim, Y. B.; Ahn, S. H.; Lee, J.-S.; Jang, C. H.; Yoon, H.; Chun, W.; Kim, G. H. Adv. Healthcare Mater. 2015, 4, 1359−1368. (17) Bertassoni, L. E.; Cardoso, J. C.; Manoharan, V.; Cristino, A. L.; Bhise, N. S.; Araujo, W. A.; Zorlutuna, P.; Vrana, N. E.; Ghaemmaghami, A. M.; Dokmeci, M. R.; Khademhosseini, A. Biofabrication 2014, 6, 024105. (18) Venugopal, J.; Ramakrishna, S. Tissue Eng. 2005, 11, 847−854. (19) O’Brien, F. J.; Harley, B. A.; Yannas, I. V.; Gibson, L. J. Biomaterials 2005, 26, 433−441. (20) Yannas, I. V.; Lee, E.; Orgill, D. P.; Skrabut, E. M.; Murphy, G. F. Proc. Natl. Acad. Sci. U. S. A. 1989, 86, 933−937. (21) Schoof, H.; Apel, J.; Heschel, I.; Rau, G. J. Biomed. Mater. Res. 2001, 58, 352−357. (22) Wahl, D. A.; Czernuszka, J. T. Eur. Cells Mater. 2006, 11, 43−56. (23) Glowacki, J.; Mizuno, S. Biopolymers 2008, 89, 338−344. (24) Ahn, S. H.; Yoon, H.; Kim, G. H.; Kim, Y. Y.; Lee, S. H.; Chun, W. Tissue Eng., Part C 2010, 16 (5), 813−820. (25) Kim, G. H.; Ahn, S. H.; Kim, Y. Y.; Cho, Y.; Chun, W. J. Mater. Chem. 2011, 21 (17), 6165−6172. (26) Suri, S.; Schmidt, C. Tissue Eng., Part A 2010, 16 (5), 1703− 1716. (27) Ghorbanian, S.; Qasaimeh, M. A.; Akbari, M.; Tamayol, A.; Juncker, D. Biomed. Microdevices 2014, 16, 387−395. (28) Lee, K. H.; Shin, S. J.; Park, Y.; Lee, S.-H. Small 2009, 5, 1264− 1268.

■

CONCLUSIONS In this study, a collagen-based cell-embedded mesh structure, manufactured using a combinational method of a cell-printing process supplemented with a core−sheath nozzle and an aerosol cross-linking process, was prepared. Using various processing conditions (pneumatic pressure in the core region and concentration of the calcium chloride solution), microsized cylindrical cell-laden struts, which consisted of a core region (cell-laden collagen) and sheath region (pure alginate), were obtained with high cell viability. To show the feasibility as a tissue engineering cell-laden structure, in vitro cellular activities including live−dead results, cell proliferation, and metabolic activity, were examined. The cell-laden structure, constituted with core−sheath regions, showed significantly higher initial cell-viability and cell proliferation than those of the control cellladen structure fabricated with an alginate-based cell printing process. Furthermore, a human ASCs-laden cell-printed structure with core−sheath region was fabricated with the same conditions, and showed significantly higher cell viability than the control. After 20 days of cell culture in a hepatogenic media, ALB and TDO2 expression were evaluated in the control and the new cell-laden structure, and the level of liverspecific gene expression in the core−sheath structure was considerably higher than in the control. Based on these results, we believe that the cell-laden core−sheath structure, fabricated using the combined process, may have significant potential as a cell-laden structure for future tissue regeneration.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.5b01764. (PDF)

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Address: Department of Biomechatronic Engineering, College of Biotechnology and Bioengineering, Sungkyunkwan University (SKKU), Suwon, South Korea. E-mail: gkimbme@skku. edu; Tel.: +82-31-290-7828. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This study was partially supported by a grant from the National Research Foundation of Korea funded by the Ministry of J

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules (29) Lee, K. H.; Seo, C. R.; Ku, J. M.; Lee, H. J.; Yoon, H.; Lee, J. H.; Chun, W.; Park, K. W.; Kim, G. H. RSC Adv. 2015, 5 (37), 29335− 29345. (30) Kim, Y. B.; Kim, G. H. J. Mater. Chem. B 2013, 1 (25), 3185− 3194. (31) Ahn, S. H.; Kim, G. H. J. Mater. Chem. B 2015, 3, 9132−9139. (32) Arvidson, S. A.; Wong, K. C.; Gorga, R. E.; Khan, S. A. Polymer 2012, 53, 791−800. (33) Cohen, D. L.; Malone, E.; Lipson, H.; Bonassar, L. J. Tissue Eng. 2006, 12, 1325−1335. (34) Haraguchi, Y.; Shimizu, T.; Yamato, M.; Okano, T. Cardiol. Res. Pract. 2011, 2011, 845170. (35) Kwak, E. A.; Ahn, S. J.; Jaworski, J. Biomacromolecules 2015, 16 (6), 1761−1770. (36) Locke, M.; Feisst, V.; Dunbar, P. R. Stem Cells 2011, 29, 404− 411. (37) Kodali, A.; Lim, T. C.; Leong, D. T.; Tong, Y. W. Macromol. Biosci. 2014, 14, 1458−1468. (38) Wang, M.; Pei, H.; Zhang, L.; Guan, L.; Zhang, R.; Jia, Y.; Li, B.; Yue, W.; Wang, Y.; Pei, X. Tissue Eng., Part C 2010, 16, 1041−1050. (39) Di Resta, G. R.; Nathan, S. S.; Manoso, M. W.; Casas-Ganem, J.; Wyatt, C.; Kubo, T.; Boland, P. J.; Athanasian, E. A.; Miodownik, J.; gorlick, R.; Healey, J. H. Ann. Biomed. Eng. 2005, 33, 1270−1280.

K

DOI: 10.1021/acs.biomac.5b01764 Biomacromolecules XXXX, XXX, XXX−XXX