The investigated system integrates cross-scale manipulation of hydrogel droplets and cells with modulable encapsulation of hydrogel microstructures. This system comprises 2 major components: electric field force-based manipulation for moving hydrogel droplets and cells and DLP for photopolymerizing hydrogel microstructures using DMD (Fig. 1A). Electric field force-based manipulation was achieved via DMF consisting of 2 plates made of indium tin oxide (ITO)-coated glass slides. The top plate features a continuous ITO coating serving as the ground electrode. The ITO coating on the bottom plate was wet etched to yield individually addressable electrodes, which were covered with a dielectric layer to provide insulation and protection. Both plates were coated with a hydrophobic layer to achieve a high contact angle, ensuring the desired fluid behavior and minimizing surface contamination. The DMF chip was assembled by adhering the top and bottom plates to double-sided tape to form a chamber. The hydrogel droplets and cells were manipulated within this chamber by applying specific electrical signals to individual addressable electrodes on the DMF, generating a localized electric field to induce EWOD or DEP. Hydrogel droplets were driven by EWOD along individually addressable electrodes, allowing for the manipulation of hydrogel droplets in various ways, including their generation, transportation, splitting, and merging on DMF. In contrast, the cells were manipulated by DEP within hydrogel droplets, which enabled adjustment of the cell density or arrangement in the hydrogel droplets. Compared to alternative hydrogel delivery technologies such as continuous microfluidics and hydrogel change cartridges, the electric field force-based manipulation system demonstrated superior flexibility, efficient hydrogel droplet manipulation [53,54], and the ability to manipulate cells within hydrogel droplets.

In the DLP system, a high-power UV beam is appropriately tuned to the desired beam diameter using beam-expanding optics. Then, this beam is propagated onto the surface of the DMD at an angle of 66°. The surface of the DMD comprises hundreds of thousands of aluminum micromirrors (1,024 × 768). When a digital mask (Fig. 1B), which is a binary image designed using PowerPoint software, is input into the DMD, each aluminum micromirror can be dynamically modulated into an ON state and an OFF state according to the corresponding black and white pixels in the digital mask. This modulation technique enables the DMD to selectively reflect the incident light beam, generating the desired light pattern. To ensure the precision of the microstructure, the light pattern is projected onto the hydrogel droplets using projection optics. When exposed to UV light, photopolymerization occurs within the hydrogel droplets, resulting in the formation of microstructures with corresponding shapes. After polymerization, the cells are immobilized and encapsulated in the hydrogel microstructures. The integrated system includes 2 electrical manipulations to drive hydrogel droplets to specific locations and adjust the density or arrangement of cells within the hydrogel droplets. Subsequently, a dynamic light pattern is used to control the photopolymerization of the hydrogel droplets, enabling the construction of programmable hydrogel microstructures with controllable geometries, components, cell densities/types, and cell patterns (Fig. 1C).

To demonstrate the flexibility of the proposed electric field force-based manipulation method, experiments were performed to manipulate hydrogel droplets and cells or microparticles. An electrical signal with an amplitude of 60 V and a frequency of 1 kHz was applied to the individually addressable electrodes, driving a hydrogel droplet to this electrode through the EWOD. Then, the amplitude and frequency of the electrical signal were changed to 15 V and 20 kHz, respectively, before the signal was applied to the electrode. This change in conditions drove the microparticles within the hydrogel droplets to be gathered by the DEP force (Movie S1). The EWOD force driving the hydrogel droplet in a parallel-plate DMF can be described as follows [55]:where ε₀ is the permittivity of vacuum, ε_D is the permittivity of the dielectric layer, L is the width of the addressable electrode, d is the thickness of the dielectric layer, and VD is the voltage of the dielectric layer.

The DEP force acting on a microparticle in hydrogel droplets can be described as follows [56]:where ε_m is the permittivity of the hydrogel droplet, r is the radius of the cell or microparticle, ∇E² is the gradient of the square of electric field strength, and Re[K(ω)] is the real part of the Clausius–Mossotti (CM) factor, which is described as follows:where ε^∗ = ε − jσ/ω, and ε and σ are the permittivity and conductivity, respectively. The subscripts p and m denote the cell or microparticle and the hydrogel droplet, respectively. In addition, ω = 2πf, where f is the frequency of applied voltage.

To further analyze the performance of EWOD manipulation for hydrogel droplets and DEP manipulation for cells or microparticles, the EWOD and DEP profiles were simulated using a 3-dimensional model in COMSOL Multiphysics (Fig. S1A and B). A corresponding cross-section of a parallel plate of DMF containing hydrogel droplets is shown in Fig. 2A. Based on this cross-section, a simplified equivalent circuit diagram of the DMF was proposed. The figure shows that the applied voltage is primarily distributed across the droplet and dielectric layer, where VL is the droplet voltage and VD is the dielectric layer voltage. According to Eqs. 1 and 2, the EWOD force is closely related to the voltage VD, and the DEP force is closely related to the electric field generated by the voltage VL. Therefore, we first simulated the voltage distribution on DMF when 3 different dielectric materials were used as dielectric layers. The resulting distributions of the voltage VD and the voltage VL are shown in Fig. 2B (Fig. S2A and B). As the amplitude of the applied voltage increases, the voltages VD and VL gradually increase. In the low-frequency range, the VD exhibits a larger voltage, whereas the VL shows a larger voltage in the high-frequency range. As shown in Eq. 1, the larger the VD is, the greater the EWOD force. However, considering the saturation of the contact angle and the dielectric strength of the dielectric layer, VD cannot be excessively large. Therefore, we selected an applied voltage of 60 V and a frequency of 1,000 Hz to analyze EWOD. Under these conditions, the voltage VD on the right electrode is 59.68 V, leading to a reduced contact angle on the right side of the hydrogel droplet. This effect generates a pressure gradient within the hydrogel droplet, driving it to move right, as depicted by the white arrow of Fig. 2C (which indicates the droplet velocity). Based on the simulation analysis, we conclude that EWOD can occur regardless of the dielectric material used, provided that the appropriate voltage amplitude and frequency are chosen to ensure the appropriate voltage VD. However, implementing DEP on DMF becomes more challenging due to the presence of a dielectric layer. This challenge arises because of the existence of the dielectric layer, which causes the frequency of the applied voltage not only to affect the CM factor (that determines the positive and negative DEP of the cells or microparticles) but also to affect the voltage VL on the hydrogel droplet (which influences the magnitude of the DEP force on the cells or microparticles). Thus, we selected 2 commonly used microparticles and cells, polystyrene microspheres, and mammalian cells and calculated their CM factors in hydrogel droplets. The results showed that the CM factor of polystyrene microspheres in hydrogel droplets remained consistently negative, while the CM factor of mammalian cells transitioned from negative to positive at kilohertz frequencies (Fig. 2D). Next, by combining the simulation results of the voltage VL with the CM factor of the microparticle and cell, the results show that when using SU-8 and Al₂O₃ as the dielectric layer, the voltage VL remains minimal before the CM factor of the mammalian cells becomes positive. However, when Ta₂O₅ was used as the dielectric layer, the voltage VL amplitude was approximately 2.4 V. Under these conditions, negative DEP was simulated for mammalian cells. The simulation results for negative DEP in mammalian cells are shown in Fig. 2E (Fig. S3A and B). The results indicated that the voltage VL was too low to induce negative DEP in mammalian cells, which was in favorable qualitative agreement with the experimental observations. Therefore, when using DEP for cell or microparticle manipulation in DMF, it is necessary to consider a suitable voltage and frequency, as well as the selection of an appropriate dielectric layer material. Based on the above simulation analysis, we can efficiently select the appropriate parameters to achieve positive and negative DEP as well as EWOD manipulation of DMF.

The hydrogel microstructures were constructed within DMF using a systematic process. First, hydrogel droplets were loaded into the chambers of DMF. These droplets were manipulated to move to their designated positions and were then exposed to a UV light pattern. The UV light pattern, projected through a highly transparent ITO-coated glass on the top plate of the DMF chip, triggered the polymerization of the hydrogel droplet, resulting in the immediate formation of the corresponding hydrogel microstructures. As a result, these hydrogel microstructures adhere to the top plate due to the superior adhesion of the hydrophobic coating 3-(trimethoxysilyl) propyl methacrylate (hydrophobic contact angle about 100°) on the top plate compared to the hydrophobic Teflon coating (hydrophobic contact angle about 115°) on the bottom plate [57]. Any residual hydrogel was carefully removed, and the hydrogel microstructures were washed in this position using a wash droplet. Next, a new hydrogel droplet was transferred to the same position using EWOD. Again, the predesigned UV light pattern was projected onto the hydrogel droplet, causing it to polymerize and form new hydrogel microstructures. This sequential process allowed for the creation of heterogeneous hydrogel microstructures with precise control of their geometry, component, and cell type distribution.

The resolution of the hydrogel microstructures was evaluated by designing specific patterns. This evaluation process began with the creation of a gradient linear microstructure (Fig. 3A), which exhibited well-defined microgrooves with an approximate width of 10 μm. Furthermore, complex-shaped hydrogel microstructures resembling a lung, fingerprint, or butterfly morphology (Fig. 3B to D) were constructed with remarkable precision, and these microstructures exhibited well-defined shapes and sharp edges. To further examine the homogeneity of the hydrogel microstructures, a pentagram-shaped array of fluorescent hydrogel microstructures was constructed by incorporating a fluorescent dye into the hydrogel droplet (Fig. 3E). Three randomly selected lines within the hydrogel microstructure array were analyzed, and the corresponding fluorescence intensity for each line was plotted (Fig. 3M). These lines exhibited similar fluorescence intensities, demonstrating the uniformity of the hydrogel microstructure array. The fabrication process allowed for precise control of the shape and material distribution of the hydrogel microstructures, as illustrated by the example of a dragon-shaped microstructure (Fig. 3F). This result highlights the versatility and accuracy of the technique in producing complex microstructures. Furthermore, various hydrogel microstructures containing fluorescent microparticles, including single microstructures (Fig. 3I), composite microstructures (Fig. 3K), single array microstructures (Fig. 3J), and composite array microstructures (Fig. 3L), were successfully constructed. The successful construction of these high-precision hydrogel microstructures with complex shapes, a regional distribution of hydrogel material, and microparticle types demonstrated the reliability of the on-chip constructed microstructures.

Previous studies have demonstrated that the UV exposure dose impacts the distribution of Young's modulus on hydrogel microstructures [58,59]. In this study, to optimize the Young's modulus of the hydrogel microstructures, the exposure time for various exposure areas was dynamically adjusted without changing the light intensity, thereby enabling the adjustment of different UV exposure doses in those areas. For example, the hexagonal star-shaped hydrogel microstructure was fabricated with varying exposure times (1.5, 3, 4.5, 6, 7.5, and 9 s) in each corner region, as shown in Fig. 3G (Movie S2). In addition, square composite hydrogel microstructures composed of PEGDA and GelMA hydrogels were exposed for selective exposure times of 2 and 4 s, as shown in Fig. 3H. After biofabrication, the Young's modulus of both the hexagonal star hydrogel microstructure and the square composite hydrogel microstructure were measured using atomic force microscopy. The Young's modulus for each area was calculated by averaging the 5 measurements (Fig. 3N and O). The distribution of Young's modulus in different regions of the hexagonal star hydrogel microstructure is shown in Fig. 3N. As the exposure time increased to 6 s, the Young's modulus exhibited a progressive increase. However, after 6 s, the Young's modulus of the hydrogel microstructure remained stable, indicating the completion of the cross-linking reaction and the cessation of any further changes caused by excess light exposure. The Young's modulus test values for the square composite microstructure comprising the PEGDA hydrogel and the GelMA hydrogel are shown in Fig. 3O. Notably, the Young's modulus of the PEGDA hydrogel microstructure and the GelMA hydrogel microstructure showed significant differences at identical exposure times, and this difference increased with increasing exposure times. These experiments highlighted that the system is capable of not only constructing various high-precision hydrogel microstructures with complex shapes, regional distributions of hydrogel materials and microparticle types but also regulating the distribution of Young's modulus within these microstructures during the fabrication process.

Under the controlled application of electrical signals, microparticles within a hydrogel droplet are manipulated using DEP. The migration of microparticles (from the left to the right side of the hydrogel droplet) is driven by DEP, as shown in Fig. 4A. Subsequently, the hydrogel droplet is split into 2 subdroplets using EWOD, allowing adjustment of the microparticle density within the hydrogel droplet (Movie S3). Similarly, the merging of 2 subdroplets also alters the microparticle density within the hydrogel droplet. The hydrogel droplet with controlled microparticle density is then cured by UV, resulting in the formation of hydrogel microstructures encapsulating microparticles with different densities. The process for adjusting the microparticle density of the polystyrene microspheres is shown in Fig. 4B. The initial density distribution started at T = 0 s, followed by gradual aggregation from left to right, which was driven by negative DEP. Finally, the density distribution was determined after the hydrogel droplet split into 2 subdroplets, as shown in Fig. 4C and D. The microparticle density of one of the subdroplets decreased in comparison to the density of the mother microdroplet, as shown in Fig. 4C. Conversely, the microparticle density of the other subdroplets increases in comparison to the density of the mother droplet, as shown in Fig. 4D. Therefore, the integration of EWOD and DEP on DMF enables efficient and noncontact adjustment of the microparticle density within hydrogel droplets.

Furthermore, shape-controllable hydrogel microstructures with various microparticle densities (Fig. 4E to K) were constructed using DLP to photopolymerize hydrogel droplets with adjusted microparticle density. In Fig. 4E to G, the gradient distribution of single fluorescent microparticles within the hydrogel microstructures is demonstrated. Specifically, as shown in Fig. 4E, the density of fluorescent microparticles within the circular hydrogel microstructure gradually increased in the radial direction (Movie S4). As shown in Fig. 4F, the hydrogel microstructure, shaped like the logo graphic of the Beijing Institute of Technology, exhibited a gradual decrease in the density of fluorescent microparticles from left to right. Finally, as shown in Fig. 4G, a planar cubic hydrogel microstructure with varying densities of fluorescent microparticles on each of its 3 surfaces was demonstrated. In Fig. 4H to J, the density gradient distribution of green and red fluorescent microparticles within the composite hydrogel microstructure was shown. Specifically, as shown in Fig. 4H, in the square composite hydrogel microstructure, the green fluorescent microparticles were located at the upper side, with their density decreasing from left to right, while the red fluorescent microparticles were located at the lower side, exhibiting a similar density trend as the green fluorescent microparticles. As shown in Fig. 4I, within the circular composite hydrogel microstructure, the green fluorescent microparticles exhibited a gradient density distribution within 3 fan-shaped regions, whereas the red fluorescent microparticles displayed a gradient density distribution within rectangular and V-shaped regions. As shown in Fig. 4J, in the circular composite hydrogel microstructure, the green fluorescent microparticles demonstrated a gradient density distribution in the upper 3 regions, while the red fluorescent microparticles demonstrated a gradient density distribution in the lower 3 regions. A hydrogel microstructure with a density gradient of mixed microparticles is shown in Fig. 4K. In this microstructure, the density of the red fluorescent microparticles increased from left to right, while that of the green fluorescent microparticles decreased from left to right. The mixed microparticle density was further quantified by normalizing the fluorescence intensity, as shown in Fig. 4L. The density of microparticles in hydrogel droplets can be adjusted through the combined operation of EWOD and DEP. Hydrogel microstructures were constructed through DLP to photopolymerize hydrogel droplets. This method enables the construction of composite hydrogel microstructures with complex shapes, a regional distribution of hydrogel material, and microparticle type/density.

The integrated system enables the regulation of microparticle density within hydrogel droplets to construct a hydrogel microstructure with controllable microparticle density. However, the microparticles within the hydrogel microstructure are gradient distributed and cannot form microparticle patterns. To address this limitation, an addressable electrode was further patterned through wet etching. This operation allowed the selective formation of microparticle patterns in specific regions using DEP. As described in the previous section, 2 hydrogel droplets with different microparticle densities were generated through the cooperative operation of EWOD and DEP. As shown in Fig. 5A and B, microparticles within the 2 hydrogel droplets gradually aggregated to form microparticle patterns under the influence of DEP (Movie S5). Notably, there was a significant difference in the size of the microparticle patterns between the hydrogel droplets. This further demonstrated that the integrated system is capable of not only regulating the microparticle density within hydrogel droplets but also forming localized microparticle patterns of different sizes.

Furthermore, hydrogel droplets containing microparticle patterns underwent photopolymerization via DLP to construct shape-controllable hydrogel microstructures encapsulating microparticle patterns (Fig. 5C to L). As shown in Fig. 5C and D, 2 square hydrogel microstructures contained microparticle patterns of different sizes while maintaining the same shape. Similarly, as shown in Fig. 5E and F, both the square and round hydrogel microstructures contained microparticle patterns of identical sizes and shapes. As shown in Fig. 5G and H, the composite hydrogel microstructures contained various microparticle patterns of the same size and shape. As shown in Fig. 5I and J, the composite hydrogel microstructure contains linear patterns of different kinds of microparticles that cross each other to form a complex pattern. Finally, hydrogel microstructures with composite attributes, including mixed shape, microparticle type, density, and pattern, was constructed (Figs. 5K and L). As shown in Fig. 5K, a hexagonal hydrogel microstructure nests with a circular hydrogel microstructure. The hexagonal microstructure contained uniformly distributed green fluorescent microparticles, while the circular hydrogel microstructure contained radial red fluorescent microparticles. As shown in Fig. 5L, a pentagram-shaped hydrogel microstructure nests with a circular hydrogel microstructure. The pentagram hydrogel microstructure contained uniformly distributed green fluorescent microparticles, while the circular hydrogel microstructure contained spherical red fluorescent microparticles. These results demonstrated the versatility of the integrated system, enabling the construction of microstructures with different shapes, microparticle densities, and various microparticle patterns.

In vitro, the construction of hydrogel microstructures with precise control of their shape, ECM, cell density, and cell pattern to replicate real tissue microstructures with high biomimicry offers immense potential for applications in the field of functional tissue engineering. For example, such composite hydrogel microstructures may be suitable for broad applications in drug modeling, drug delivery, and tissue repair. To examine cell growth within the hydrogel microstructure, we incubated the octagonal hydrogel microstructures encapsulating NIH3T3 cells in a biological incubator. Cell viability was assessed by using live/dead staining after 15 d of culture (Fig. 6A). As shown in Fig. 6B, cell viability was approximately 70% on the first day of culture. Over time, the cell viability steadily increased, indicating that the hydrogel microstructures constructed through the system did not adversely affect cell activity. As shown in Fig. 6C, the HepG2 cell density within the hydrogel droplet was adjusted. Subsequently, the hydrogel droplet was mixed with another hydrogel droplet containing NIH3T3 cells and then photopolymerized to construct hydrogel microstructures that incorporated both cell types. Different initial densities of HepG2 cells were incorporated into these microstructures. For these hydrogel microstructures, NIH3T3 cells were labeled with green fluorescence, and HepG2 cells were labeled with red fluorescence. Bright-field and fluorescence images taken on day 7 of hydrogel microstructure culture are shown in Fig. 6D. Furthermore, the quantitative analysis of the proliferation rate of the 2 cell types within the hydrogel microstructure is shown in Fig. 6H. Notably, HepG2 cells exhibited a slower proliferation rate than NIH3T3 cells. Therefore, when the cells are cultured together, regulating their initial density is crucial for achieving the desired cell distribution. We further constructed hydrogel microstructures with a gradient of NIH3T3 cell densities and patterned distribution of NIH3T3 and HepG2 cell aggregates, as shown in Fig. 6E (a to c). These hydrogel microstructures underscore the functionality of our biofabrication technology. Additionally, 2 different biomimicry microstructure models of the hepatic lobule were constructed, as shown in Fig. 6F and G. The first model involved coculturing NIH3T3 cells and HepG2 cells within hydrogel microstructures. In contrast, the second model included NIH3T3 cells and the ordered arrangement of HepG2 cells within hydrogel microstructures. Urea secretion was measured in these biomimicry microstructure models of the hepatic lobule during 2 cultivation cycles. As shown in Fig. 6I, the second model exhibited significantly greater urea secretion than the first model. This difference indicates that the order of arrangement of cell states within hydrogel microstructures affects the functional expression of microstructure models. Thus, these microstructures provide a platform for constructing biomimetic biological microstructures that enable study of cell behavior and interactions within specific microenvironments, facilitating a deeper understanding of cell signaling mechanisms. The potential applications of these microstructures are proposed to drive further development and innovation in related fields, particularly in medicine and related disciplines.

The DMF device consists of core components, including a DMF chip, a microcontroller unit, a signal generator (33600A, KEYSIGHT), a voltage amplifier (ATA-2021B, Aigtek), and a high-voltage ac switching matrix. The signal generator generates an ac signal, which is then amplified by the voltage amplifier before being sent to the high-voltage ac switching matrix. The switching matrix is connected to the electrodes of the DMF. Coordinated with instructions from the host computer, the microcontroller generates precise digital signals that control the switching mechanism within the high-voltage ac switching matrix, enabling seamless activation and deactivation of the electrodes on the DMF. By applying appropriate electrical signals to these electrodes, we performed targeted EWOD and DEP operations to achieve controlled manipulation of hydrogel droplets and microparticles.

Dynamic optical projection lithography consists of several key components, including a UV light source (5 W, 405 nm; Shanghai Runcast Electronic Technology Co., Shanghai, China), beam expander optics, DMD (V-7000VIS, TX, USA), projection optics and a charge-coupled device. The emitted UV light beam is directed through the beam expander optics to adjust its diameter. This modified light beam is then projected to the DMD, which is responsible for reflecting a patterned light beam aligned with the input design mask. The patterned light beam, now refined, travels through the projection optics onto a hydrogel droplet, resulting in its controlled photopolymerization process.

A PEGDA hydrogel was prepared by dissolving PEGDA (Mn 3400; Laysan Bio, USA), acrylate-PEG-RGDS (JenKem Tech, USA) and pluronic F127 (Sigma-Aldrich, St. Louis, MO USA) in low-conductive buffer (LCB) (10 mM Hepes, 0.1 mM CaCl₂, 59 mM D-glucose, and 236 mM source, pH 7.35). A 50% (w/v) solution of photoinitiator was prepared by dissolving 2-hydroxy-1-(4-(hydroxyethoxy)phenyl)-2-methyl-1-propanone (Irgacure 2959, BASFSE, Germany) in dimethyl sulfoxide (Fisher Scientific, USA) and adding it to the PEGDA hydrogel to be used as a photoinitiator. The final concentrations of the components in the PEGDA hydrogel were 15% w/v for PEGAD, 0.5% w/v for photoinitiator, 0.1 w/v for pluronic F127, and 5 mM RGDS in the LCB.

A GelMA hydrogel was prepared by dissolving GelMA prepolymer (Suzhou Yongqinquan Intelligent Equipment Co., Suzhou, China), pluronic F127 (Sigma-Aldrich), and lithium phenyl-2,4,6-trimethyl-benzoylphosphinate (LAP, Allevi, USA) into the LCB. The final concentrations of the components in the GelMA hydrogel were 5% w/v for GelMA, 0.5% w/v for LAP, and 0.1% for pluronic F127 in the LCB.

HepG2 (human hepatocellular carcinoma) and NIH3T3 (mouse fibroblast) cells were purchased from the American Type Culture Collection (Manassas, VA, USA). HepG2 and NIH3T3 cells were cultured separately in Dulbecco's modified Eagle's medium (HyClone, Logan, UT, USA) supplemented with 10% (v/v) fetal bovine serum (Gibco, Gaithersburg, MD, USA) and 1% (v/v) penicillin–streptomycin solution (Solarbio, Beijing, China). The samples were stored in an incubator supplemented with 5% CO₂ at 37 °C, and the medium was changed every 2 to 3 d. Before encapsulation, HepG2 and NIH3T3 cells were collected separately and resuspended in LCB. To distinguish the distribution of different cell types in the hydrogel microstructure, HepG2 cells were stained red, and NIH3T3 cells were stained green by PKH26 and PKH67 (Sigma-Aldrich).

Live/dead staining assays of the cell encapsulation within the hydrogel microstructures were performed by labeling with calcein acetoxymethyl ester (2 μg/ml) and propidium iodide (3 μg/ml) fluorescent dyes (Molecular Probes, Eugene, OR, USA). Cellular viability within the hydrogel microstructures was assessed daily using ImageJ software to analyze the areas of red fluorescence (dead cells) and green fluorescence (live cells) in each image. Cellular viability was quantified by calculating the ratio of the area of green fluorescence relative to the total area of fluorescence.

For urea secretion, the amount of urea secreted into the medium was measured using a urea assay kit (Sigma-Aldrich). Before the assay, the constructed microstructures were washed with Dulbecco's modified Eagle's medium and placed in fresh medium. After 24 h, the medium was collected and temporarily stored at −20 °C. After all the control groups were sampled, the solutions were gradually thawed to room temperature for the assays. The urea concentration was analyzed using a urea assay kit (Sigma-Aldrich). The absorbance was read using a microplate reader.