The authors have declared that no competing interests exist.
The creation of engineered 3D microtissues has attracted prodigious interest because of the fact that this microtissue structure is able to mimic in vivo environments. Such microtissues can be applied extensively in the fields of regenerative medicine and tissue engineering, as well as in drug and toxicity screening. Here, we develop a novel method of fabricating a large number of dense honeycomb concave microwells via surface tension-mediated self-construction. More specifically, in order to control the curvature and shape of the concavity in a precise and reproducible manner, a custom-made jig system was designed and fabricated. By applying a pre-set force using the jig system, the shape of the honeycomb concave well was precisely and uniformly controlled, despite the fact that wells were densely packed. The thin wall between the honeycomb wells enables the minimization of cell loss during the cell-seeding process. To evaluate the performance of the honeycomb microwell array, rat hepatocytes were seeded, and spheroids were successfully formed with uniform shape and size. Liver-specific functions such as albumin secretion and cytochrome P450 were subsequently analyzed. The proposed method of fabricating honeycomb concave wells is cost-effective, simple, and reproducible. The honeycomb well array can produce multiple spheroids with minimal cell loss, and can lead to significant contributions in tissue engineering and organ regeneration.
Multicellular organisms consist of cells organized in three-dimensional (3D) environment in which cells continuously interact with neighboring cells. And cells cultured in 3D environment shows more in vivo cellular behaviors than that of 2D environment such as differentiation, proliferation, and gene expression [
Several 3D cell culture methods have been developed for engineering 3D microtissues, including hanging drop [
Recent progress in microfabrication technology has facilitated the development of diverse microscale platforms for engineering 3D tissues using simple and cost-effective processes that employ a wide range of materials. The representative microplatform is the microwell array, and various types of cells–including stem cells, cancer cells, liver cells, pancreas cells, and brain cells–have been used to create different tissue models [
Another critical issue involves the seeding of cells with minimum cell loss, as certain cells are expensive and difficult to access. Furthermore, obtaining the maximum production of 3D tissues within a limited area is important from the perspective of cost and labor. For this purpose, the thickness of the wall between microwells should be small, and the position of arrayed microwells should be as compact as possible. However, it is more challenging to fabricate such arrayed microwell structures with concave bottoms.
We have previously developed two simple methods for fabricating concave bottoms. According to one method, a thin elastic membrane is deflected by applying vacuum pressure, and the deflected membrane is replicated using UV-curable material [
In this paper, we develop a custom-made machine for fabricating the homogeneous shape of the concave microwell by applying uniform force and raking the prepolymer with uniform speed. The force on the PDMS microwells is automatically controllable, and the curvature of the concave structure can be modulated by varying the amount of force. Moreover, the uniform speed applied by an automated linear stage enhances the uniformity of the well shape. With the proposed device, honeycomb-structured concave microwells with thin walls were fabricated in order to minimize cell loss by decreasing the dead area of walls between microwells. As the thickness of walls decreases, most seeded cells fall into honeycomb concave microwells because the top area of the wall is narrow, thus making it possible to effectively seed cells with minimal cell loss and without a laborious cell-washing process. This honeycomb structure therefore enables the increase of spheroid density; additionally, the consumption of media and diverse biomolecules could be reduced. The performance of this system was verified by preparing and seeding primary rat hepatocytes and demonstrating the successful formation of spheroids. The creation of well-defined spheroids with the desired size and properties (e.g., co-cultured cell populations) is acutely important for studies concerning liver physiology, drug efficacy and the screening of toxicity, and the regeneration of liver function [
Arrayed PDMS honeycomb microwells with flat bottoms and various diameters (300–1000 μm) were fabricated using a conventional soft lithography process [
(a) Poly(dimethylsiloxane) (PDMS) honeycomb structures are fabricated using standard lithography. (b) A PDMS prepolymer is poured onto a PDMS base chip to fill the microwell, (c) and is then raked out using a flat plate with trimmed edges. (d, e) Surface tension-mediated honeycomb concave microwells are thermally cured on a hot plate. (f) A custom-designed jig system is used to fabricate honeycomb concave microwells. (g) Constant force is applied to the chip on the linear stage. (h) The PDMS prepolymer is raked out from the chip as the motorized linear stage moves in one direction.
A custom-made jig system was developed for fabricating homogeneous concave microwells using the two-step process of application of constant force and raking to a prepolymer-filled chip (
SD (Sprague Dawwley) male rats (8 weeks) (KOATECH, Republic of Korea) were sacrificed by CO2 inhalation for the isolation of hepatocytes and hepatic stellate cells. Primary hepatocytes were isolated from 8-week-old, male Sprague-Dawley rats (DBL, Seoul, Republic of Korea) using a two-step collagenase perfusion procedure [
Hepatocytes were seeded and spheroids were self-aggregated in PDMS-based concave microwells. Concave microwell array chips were sterilized by being autoclaved and dried in an oven for 1 h. In order to prevent cell attachment, we coated the concave PDMS chip with 3% (w/v) bovine serum albumin (BSA) overnight in an incubator, after which the chip was rinsed with phosphate-buffered solution (PBS). A suspension of primary hepatocytes (400 μL with 4 × 105 cells/mL) was directly seeded onto the concave microwells. The cells were evenly docked within the concave microwells.
Cell loss during the seeding process was measured by counting the cells in the external region of the microwells. After seeding hepatocytes in each group of microwell, we gathered the undocked cells by pipetting and counting the number of cells retrieved. Cell loss ratios were calculated as the number of cells retrieved from pipetting divided by the initial number of seeded cells, and the ratios of each group (i.e., the circular and honeycomb arrays) were compared. After 7 days of culturing, we measured the number of cells in each spheroid of each group. We counted the number of spheroids and treated them with a 1:1 mixture of TrypLE Express (gibco) and 1X HBSS (gibco) for 5 min at 37°C, and then dissociated them into single cells by vortexing. We counted the number of single cells and calculated the number of cells in a single spheroid.
Spheroids were retrieved from each type of microwell and fixed by immersing in 4% paraformaldehyde (PFA) for 30 min at 4°C and then incubating in 0.1% Triton X-100 in PBS for 15 min at room temperature. After washing with PBS containing 0.1% BSA, the spheroids were incubated with 3% BSA at room temperature for 30 min. After aspirating the BSA solution, the spheroids were washed again and incubated overnight at 4°C with mouse anti-albumin antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA). At this stage, the spheroids were washed with 0.1% BSA and then incubated at 4°C for 90 min with Alexa Fluor 488-conjugated anti-mouse IgG (Invitrogen) secondary antibodies. Other groups of spheroids were prepared as described above and incubated with rabbit anti-cytochrome P450 reductase (Abcam, UK) primary antibodies. These spheroids were washed and incubated with Alexa Fluor 488-conjugated anti-rabbit IgG (Invitrogen) or Alexa Fluor 568-phalloidin (Invitrogen) for F-actin staining, as appropriate. All spheroids were then incubated with DAPI (4',6-diamidino-2-phenylindole) for 5 min at room temperature before being imaged using a confocal microscope (Olympus, Japan).
The viability of hepatocytes cultured on the three types of concave microwells was analyzed using a Live/Dead assay (Invitrogen, CA), as described by the manufacturer. 5 mL of primary hepatocyte medium (PHM) containing 2 μL of calcein-AM solution and 5 μL of ethidium homodimer-1 solution were added to each concave microwell, and then incubated at 37°C for 1 h. Stained hepatocytes were analyzed under a fluorescence microscope and confocal microscope (Olympus, Japan).
Albumin secretion was analyzed using a rat albumin enzyme-linked immunosorbent assay (ELISA) kit (KOMA Biotech, Seoul, Korea) as described by the manufacturer. 300 μL of culture supernatant was collected daily and replaced with 300 μL of fresh DMEM medium for 1 week. Supernatant samples and controls (fresh medium) were added to pre-treated 96-well plates, and absorbance was measured at 450 nm using a microplate reader. Triplicate readings were averaged, and the values obtained by subtracting the converted concentration of the fresh medium from that of the samples were considered to be the concentration of albumin secreted from the spheroids in each microwell.
Hepatocyte spheroids formed in circular concave microwells and honeycomb concave microwells were observed under a scanning electron microscope (SEM, JEOL Ltd, Tokyo, Japan). The shape of hepatocyte spheroids was analyzed by fixing the spheroids with 2.5% glutaraldehyde in deionized water (DW) for 1 h, and then washing with DW. In a secondary fixation step, the spheroids were immersed in 1% osmium tetroxide in DW for 1 h and dehydrated using a graded ethanol series (25%, 50%, 75%, 95%, and 100%). After dehydration, the spheroids were washed in tert-butanol three times for 30 min each at room temperature and frozen at −70°C. Hepatocyte spheroids were freeze-dried until the tert-butanol evaporated, and were then mounted onto a specimen stub with graphite paste, coated with palladium alloy, and observed with an SEM.
Surface tension-mediated honeycomb concave microwell arrays were successfully fabricated using a custom-made jig system.
(a) Schematic diagram of a honeycomb concave microwell array. W: width; D: diameter; T: wall thickness. (b) A picture of various sizes of honeycomb concave microwell chambers. (c) Pictures of the jig system pressing a chip. Scanning electron microscope (SEM) images of wells formed at a force of (d) 3 N and (e) 7 N showing top views and cross-sectional side views of microwells. Scale bars: 500 μm.
Curvatures of honeycomb concave microwells formed by applying various forces to the PDMS prepolymer. The black dotted lines and black triangles show the original microwell before pouring the PDMS prepolymer. The blue-dotted circle depicts the radius of the honeycomb concave microwell. Scale bars: 200 μm.
(a) Schematic depiction of side and top views of honeycomb concave microwells. (b) Relationship between well depth and radius of curvature of honeycomb concave microwells.
To produce homogeneous hepatocyte spheroids, we used two types of chips: circular concave microwells [
(a, e) Scanning electron microscopy (SEM) images showing the smooth bottom profile of concave microwells 1 and 2. (b, f) Images of hepatocyte spheroids formed in concave microwells 1 and 2. (c, g) Merged images of Calcein AM- and Ethidium homodimer-1-stained spheroids. Calcein AM-stained hepatocyte spheroids (green) cultured for 3 days showing cell viability in concave microwells 1 and 2. Ethidium homodimer-1 fluorescence images (red), which indicate dead cells, are weakly appeared on the images. Scale bars: 500 μm. (d, h) SEM images of spheroids showing formation of tight cell-to-cell junctions in spheroids (black arrows) formed in concave microwells 1 and 2. Confocal microscopic images of cell viability in circular concave microwells (i) and honeycomb concave microwells (j). Scale bars: 250 μm.
To determine the effect of each culture condition on the size of the spheroids, we analyzed the diameters of hepatocyte spheroids cultured for 7 days using Image J software (
(a) Spheroid formation in two types of concave microwells. Scale bars: 500 μm. (b) Size distribution of hepatocyte spheroids in circular and honeycomb concave microwells.
In order to investigate the relationship between the shape of the microwell and the corresponding cell loss, we performed a cell-loss test. Cells in the external region of each microwell were gathered by pipetting, and we then counted the number of retrieved cells. A high percentage (65.3%) of cells was lost in circular concave microwells because these arrays have a smaller number of microwells and a larger amount of dead space (
Well types | Number of wells | Flat area in concave microwell chamber (%) | Cell loss (%) |
---|---|---|---|
Sparse concave microwell | 100 | 80.4 | 65.3 |
Dense concave microwell | 196 | 61.5 | 26 |
Honeycomb concave microwell | 196 | 57.6 | 22 |
To analyze the liver-specific functions of hepatocyte spheroids in each type of microwell, we performed immunostaining for albumin, CYP450, and F-actin, and assessed the albumin secretion using ELISA. The liver-specific function test was performed for each concave microwell. As shown in
(a) Analysis of metabolic functions of spheroids in the two types of microwells, as measured by the secretion of albumin. (b) Immunostaining of hepatocyte spheroids for albumin (primary hepatocytes; green in first row), DAPI (nuclei; blue), cytochrome P450 activity (CYP450; green in second row) and microfilaments (F-actin; red). (c) Quantified analysis of hepatocytes function test. Scale bars: 300 μm.
We fabricated a deep honeycomb concave microwell array by exploiting the viscoelastic properties and surface tension of the PDMS prepolymer. Using a custom-developed jig system, we demonstrated the successful and reproducible formation of uniformly sized concave microwells. In our previous study, we fabricated concave microwells for the production of embryoid bodies by raking out the PDMS prepolymer using glass slides [
Upon using the concave honeycomb wells, uniformly shaped and sized spheroids were successfully fabricated from seeded rat hepatocytes. Cell loss during cell seeding was minimized, and the number of spheroids within a certain area was maximized by employing a compact honeycomb structure. A microwell depth of 650 μm was sufficiently deep to avoid the escape of spheroids from their original position during media change. As expected, cell loss during cell seeding in the circular concave microwell was high (>65%), as shown in
The function of hepatocyte spheroids was evaluated by assaying albumin secretion over 6 days and immunostaining for CYP450 reductase as an indicator of enzymatic activity. All spheroids in the two types of microwells secreted albumin continuously and showed a similar degree of enzymatic activity. However, the honeycomb concave microwells were highly effective in producing large numbers of functional spheroids upon using the same seeding and culturing procedures, as evidenced by the results of cell-loss tests and albumin ELISAs. On the basis of these results, we conclude that our honeycomb concave microwell model could provide an appropriate platform for producing uniformly sized spheroids with minimal cell loss, as well as high spheroid density and viability.
In this paper, we have developed a simple and cost-effective method for fabricating PDMS-based honeycomb concave microwell arrays using a jig system that enables control over the radius of curvature and well depth, thereby producing uniformly sized hepatocyte spheroids. The densely packed honeycomb concave microwells greatly increased the cell seeding efficiency. This system could be used for other types of agglomerate cells to form uniform-sized 3D spheroids with minimum cell loss. Additionally, the proposed method for spheroid production has extensive potential applications in tissue engineering, organ regeneration, the study of organs and diseases, and the screening of drug and toxicity.
Top view and side view of (a) 300, (b) 400, (c) 500, (d) 700, and (e) 1000 μm diameter wells. Side-view images of different chip sizes: (f) 10×10 mm2, (g) 15×15 mm2, and (h) 20×20 mm2. Scale bars: 300 μm.
(TIF)
(TIF)