Enhancement of properties of a cell-laden GelMA hydrogel-based bioink via calcium phosphate phase transition

To improve the properties of the hydrogel-based bioinks, a calcium phosphate phase transition was applied, and the products were examined. We successfully enhanced the mechanical properties of the hydrogels by adding small amounts (< 0.5 wt%) of alpha-tricalcium phosphate (α-TCP) to photo-crosslinkable gelatin methacrylate (GelMA). As a result of the hydrolyzing calcium phosphate phase transition involving α-TCP, which proceeded for 36 h in the cell culture medium, calcium-deficient hydroxyapatite was produced. Approximately 18 times the compressive modulus was achieved for GelMA with 0.5 wt% α-TCP (20.96 kPa) compared with pure GelMA (1.18 kPa). Although cell proliferation decreased during the early stages of cultivation, both osteogenic differentiation and mineralization activities increased dramatically when the calcium phosphate phase transition was performed with 0.25 wt% α-TCP. The addition of α-TCP improved the printability and fidelity of GelMA, as well as the structural stability and compressive modulus (approximately six times higher) after three weeks of culturing. Therefore, we anticipate that the application of calcium phosphate phase transition to hydrogels may have the potential for hard tissue regeneration.


Introduction
The demand for tissue regeneration and organ replacement technologies is increasing steadily worldwide.Recently, three-dimensional (3D) bioprinting has been developed to replace damaged tissue or regenerate healthy tissue to produce functional tissue [1][2][3][4].The ability to create customized designs for patients and enable high-precision, low-cost, and ondemand fabrication of complex structures in a short amount of time are beneficial for developing 3D bioprinting in the biomedical industry [5,6].
In such 3D bioprinting processes, bioink is the main component.Bioinks that integrate biomaterials-live cells in various forms (e.g.single cells, aggregated cells) and bioactive molecules (e.g.growth factors, DNA, miRNA)-have been widely used for 3D bioprinting in tissue engineering [7][8][9][10].Recently, a range of bioink formulations containing a combination of biomaterials or a solution of biomaterials that encapsulate appropriate cell types have been tested for hydrogel production [8,11].
In numerous tissue engineering fields, hydrogelbased bioinks have been extensively studied for the production of cell-laden scaffolds because of their favorable cell anchoring and metabolic activities [10,[12][13][14].However, there are a few issues to consider when using hydrogel-based bioinks for hard tissue engineering because of the swelling behavior of hydrogel bioinks and the lack of long-term fidelity and stability of the printed structures [15][16][17].The swelling behavior causes structural fragmentation owing to ongoing internal mechanical loads and delayed tissue regeneration owing to high tension [18,19].Hence, the swelling and subsequent lack of mechanical properties means that it is not possible to guarantee the long-term subaqueous fidelity of the constructs [20,21].Moreover, it has been found that encapsulated cells are frequently not viable in long-term studies [20,22,23].To overcome these obstacles, it is vital to develop a bioink that can achieve long-term subaqueous fidelity and high cell viability by enhancing the mechanical strength of the product.
There are various methods for fabricating strong and tough hydrogels, such as fiber/fabric reinforcement, aligned microstructures, self-reinforced composite materials, and micro/nanoparticle additives [23][24][25].The addition of inorganic fillers (e.g.inorganic nanoparticles, glasses, salts, and bioceramics) has been investigated in various studies to improve the strength of hydrogel composites [26][27][28].Interestingly, improvements in the mechanical properties resulting from the inclusion of additives have been found to be related to the quantity rather than type of additive.Therefore, the quantity of any additive included in a bioink should be carefully determined to ensure that it enhances both the viability of the encapsulated cells and mechanical strength of the product [29][30][31][32].Furthermore, the co-printing system has been investigated to improve the mechanical properties of the hydrogel as well as the possibility of printing cell-laden structure with appreciated cell response under the usage of biphasic bioinks as cell-laden biopolymer bioinks and self-setting, calcium phosphate-based bone cement [33][34][35].
One such additive, alpha-tricalcium phosphate (α-TCP), has been selectively studied for its ability to enhance the mechanical properties of hydrogelbased cell-laden bioinks via bone-cement chemistry.α-TCP is a calcium phosphate (CaP) that has attracted considerable attention because of its many beneficial characteristics, including its high biocompatibility.In addition, the hydration reaction of α-TCP occurs at physiological pH, and the resultant calciumdeficient hydroxyapatite (CDHA) precipitant exhibits a crystal phase that is similar to that of human bone and favorable bioactivity and mechanical properties [36][37][38][39].Therefore, α-TCP was selected for the present study.
Regarding the matrix, gelatin methacrylate (GelMA) was selected for this study.GelMA is a widely used hydrogel for bioinks because of its outstanding effects on biological activities, such as cell elongation and adhesion, which are the result of its naturally occurring cell-adhesive motifs (i.e.RGD peptide) [40][41][42][43].Additionally, it is highly biocompatible with a viscosity that is comparable to water at T < T sol−gel (i.e.T sol−gel at 30 • C) [42].To avoid structural collapse, GelMA possesses adequate rheological properties that permit extrusion during layer-bylayer deposition and shape retention after fabrication.For long-term form preservation and stability at physiological temperatures, the printed structure must also be treated with a crosslinker, which may impart toxicity [44,45].
In this study, bioink materials with enhanced mechanical properties suitable for hard tissue regeneration were developed by applying a CaP phase transition to a photocrosslinkable GelMA hydrogel, as shown in figure 1.The amount of α-TCP in the GelMA hydrogel was carefully controlled to ensure that the calcium phosphate phase transition and the environment for the cells were optimized.The effects of phase transition on the properties of the resulting hydrogels were investigated.The phase transition of optimized amounts of α-TCP and its mechanical properties were investigated.The effects of calcium phosphate phase transition and loaded α-TCP (phase transformed to CDHA after a sufficient reaction time) on the viability and activity of the encapsulated cells were investigated.After the effects of loaded α-TCP on the hydrogel were confirmed, the range of printing conditions for the fabrication and osteogenesis of the fabricated scaffolds were also investigated.

Fabrication of α-TCP/GelMA hydrogels and cell culture
To fabricate the cell-laden hydrogels, cells harvested from a mouse calvaria-derived preosteoblast cell line (MC3T3-E1, American Type Culture Collection, USA) were added to GelMA solutions containing different amounts of α-TCP homogeneously at 5 × 10 6 cells ml −1 .The resultant mixtures were stored at 4 • C for 5 min to allow the GelMA to solidify before hydrogel disc fabrication because of the viscosity of GelMA, which has a water-like low viscosity: T < T sol−gel (i.e.T sol−gel at 30 • C) [42,44].The mixed solutions were subjected to UV light irradiation (365 nm) for 1 min at room temperature (RT) for photocrosslinking and disc fabrication (1 mm height × 8 mm diameter).After fabrication, the hydrogel discs (n = 4) were transferred to 24-well plates for cell culture.
Cell-free hydrogels (1 mm height × 8 mm diameter) were prepared under the same conditions, but cells were not added.They were also immersed in osteogenic differentiation medium to allow the calcium phosphate phase transition to proceed.
The water content was calculated from the weights of the water-equilibrated (W w ) and dried (W d ) hydrogels.Briefly, each hydrogel disc was immersed in 1 ml of cell culture medium at 37 • C for 0 h and 36 h.At specified time points, the water-soaked hydrogel discs were removed from the medium, and their surfaces were gently wiped.Each sample was weighed until equilibrium weight was reached.All the samples were then freeze-dried and weighed.The swelling percentage was calculated using the following equation: To measure of the compressive modulus of the hydrogel, a rotational rheometer (DHR-1, TA Instruments, USA) was used with a flat parallel plate (8 mm, crosshatched) at 25 • C. Each sample was subjected to compressive loading at a constant linear rate of 5.0 µm s −1 .The load-displacement data were transformed into stress-strain plots.The slope of a linear fit for the strain range of 0.05-0.15was used as an acquisition of the compressive modulus.

Evaluation of the effect of the hydrogels on cellular activity and osteogenic differentiation
All the samples were cultured in osteogenic differentiation medium immediately after fabrication.At each time point (0 and 36 h), the culture medium was removed and replaced with 1 ml of cell counting kit-8 (CCK-8) solution (10% v/v in culture medium).The samples were incubated for 2 h at 37 • C, and 200 µl of supernatant was decanted to measure absorbance at 460 nm to determine cell viability.
At predetermined time points (1, 2, and 3 weeks), 1 ml of cell lysis buffer (CelLyticTM M, Sigma, USA) was added to each cell-laden hydrogel disc sample and the samples were homogenized.DNA was detected using a PicoGreen assay kit (Thermo Fisher Scientific, USA) following the manufacturer's protocol.To determine alkaline phosphatase (ALP) activity, p-nitrophenyl phosphate (pNPP) was used as the substrate.Absorbance was measured at 405 nm using a spectrophotometer.A commercial calcium assay kit (Pointe Scientific.Inc., USA) was used for the quantitative determination of calcium ions, which were determined by measuring the absorbance at 570 nm.In detail, for the calcium ion quantification, the cellfree hydrogel in each group which contained different amounts of α-TCP (0, 0.125, 0.25, 0.5 wt%) was also used for negative control.Then each time point, the amounts of quantified calcium ions of negative control groups were subtracted from the cell-laden hydrogel.

Fabrication of cell-laden scaffolds and evaluation of cell viability
Cell-laden α-TCP/GelMA solutions containing 0 or 0.25 wt% α-TCP were stored at 4 • C for 10 min.Then, uncrosslinked hydrogel scaffolds were fabricated using a customized three-axis extrusion-based printer with a nozzle (inner diameter = 390 µm) at the following temperatures: syringe temperature, 15 • C; plate temperature, 10 • C. The pneumatic pressure applied was 100 kPa.The feed rate of the printing system was set at 100 mm min −1 .The printed scaffolds of each group were subjected to UV light irradiation (365 nm) for 1 min at RT for photocrosslinking.The scaffolds were then cultured in 24-well plates in osteogenic differentiation medium.
The viability of the cell-laden α-TCP/GelMA scaffolds was determined using a live/dead viability/cytotoxicity kit for mammalian cells (Invitrogen, USA) following the manufacturer's protocol.

Alizarin red S staining
After 3 weeks of culture, cell-laden α-TCP/GelMA scaffolds containing 0 or 0.25 wt% α-TCP were fixed in 4% formaldehyde, dehydrated via a graded series of ethanol, and embedded in paraffin blocks.The sectioned samples were then deparaffinized, hydrated in a graded series of ethanol, stained with 2% (w/v) Alizarin Red S solution for 10 min, and washed with tap water to remove excess dye.Light microscopic images were taken using an optical microscope (IX71; Olympus, Japan) at 10× magnification.

Statistical analysis
Data were analyzed using one-way analysis of variance (ANOVA).Tukey simultaneous tests were used to identify differences between individual hydrogel groups using Origin software.Each cell activity experiment was repeated at least three times with four different samples in each group, and a p value < 0.05 was considered statistically significant.

Characterization of α-TCP/GelMA hydrogels formed during the calcium phosphate phase transition
The characteristics of different α-TCP/GelMA hydrogels formed during the phase transition were investigated using cell-free hydrogels containing different amounts of α-TCP.The calcium phosphate phase transition took place when the mixtures were immersed in cell culture medium for 36 h at 37 • C. The reaction was performed under physiological conditions (i.e.neutral pH and temperature < 37 • C), which are suitable for biofunctional material.The initial hydrolysis reaction for the phase conversion of α-TCP to CDHA is as follows [37,46,47]: During the calcium phosphate phase transition, α-TCP powder transitions to CDHA, which has a crystalline phase similar to that of the human  The water content of each hydrogel sample was measured to assess swelling, which can be affected by α-TCP or CDHA, at 0 h (before CaP phase transition) and 36 h (after CaP phase transition).Before calcium phosphate phase transition, each group of hydrogel samples showed a similar swelling ratio, regardless of whether they included α-TCP, as shown in figure 3(a).However, after calcium phosphate phase transition, the swelling ratio decreased at a rate that was inversely proportional to the amount of CDHA (which transitioned from α-TCP) in the hydrogel, as shown in figure 3(b).
The mechanical properties (compressive modulus) of each group of hydrogels were investigated to study the effect of α-TCP addition.Samples from each hydrogel group (containing different amounts of α-TCP) were subjected to the same conditions as those used to examine the calcium phosphate phase transition process.As shown in figure 4(a), before the calcium phosphate phase transition occurred (0 h), there were no significant differences in the compressive modulus of the hydrogel groups (0 wt% α-TCP:1.18kPa, 0.125 wt% α-TCP:1.69kPa, 0.25 wt% α-TCP:1.63 kPa, 0.5 wt% α-TCP:1.60 kPa).During the calcium phosphate phase transition, there were no significant differences in the compressive modulus of the hydrogel groups that contained 0 wt% α-TCP and 0.125 wt% α-TCP; however, the compressive modulus of the 0.25 wt% and 0.5 wt% α-TCP groups was increased throughout the entire period, as shown in figure 4(b).At the completion of the reaction period (36 h), the 0.25 and 0.5 wt% α-TCP groups showed considerable increases in their compressive modulus, which can be observed by comparing the data presented in figures 4(a) and (c).Specifically, the compressive modulus of the 0.25 wt% α-TCP group was 14.16 kPa (seven times higher than before calcium phosphate phase transition), and that of the 0.5 wt% α-TCP group was 20.96 kPa (about ten times higher than before calcium phosphate phase transition).

Viability of encapsulated cells in α-TCP/GelMA hydrogels
During the calcium phosphate phase transition, calcium ions are eluted, and the pH of the reactant solution changes because of the phase conversion of α-TCP to CDHA [37].The resultant increase in extracellular calcium ion concentration in the hydrogel causes an influx of calcium ions from the cytoplasm to the mitochondria and nuclei of the encapsulated cells.In mitochondria, calcium accelerates and disrupts normal metabolism, leading to cell death through apoptosis [50,51].Therefore, the quantity of α-TCP included in the hydrogel formulation could negatively affect encapsulated cells.
For this reason, the viability of cells encapsulated in various α-TCP/GelMA hydrogel samples was investigated using a cell counting kit (CCK-8).As shown in figure 5(a), before calcium phosphate phase transition (0 h; immediately after fabrication), there were no significant differences in viability between the hydrogel groups, indicating that there was no critical cellular damage caused by loading.However, after calcium phosphate phase transition (i.e.immersion in cell culture medium for 36 h at 37 • C), the hydrogel groups that contained α-TCP, regardless of the amount, showed a decrease in cell viability of approximately 91% compared with the hydrogel without α-TCP.To further investigate the effect of calcium phosphate phase transition on the viability and proliferation of encapsulated cells, DNA was quantified and the results are shown in figure 5(b).The encapsulated cells in every hydrogel group proliferated during the first week of culture.Even when there was a small decrease in cell viability after calcium phosphate phase transition, the encapsulated cells in the hydrogel recovered.Specifically, the cells encapsulated with 0.25 and 0.5 wt% α-TCP recovered within one week.However, compared with the 0 and 0.125 wt% α-TCP hydrogel groups, the groups containing larger amounts of α-TCP exhibited lower quantities of DNA for the entire culture period.After three weeks, the 0 and 0.125 wt % α-TCP groups produced the highest DNA quantities, as follows: 0 wt% α-TCP = 1.55 ± 0.11 µg ml −1 , 0.125 wt% α-TCP = 1.38 ± 0.11 µg ml −1 , 0.25 wt% α-TCP = 0.81 ± 0.07 µg ml −1 , and 0.5 wt% α-TCP = 0.55 ± 0.06 µg ml −1 .
Overall, after 3 weeks, all groups except the 0.5 wt% α-TCP group showed an increase in DNA quantity compared to immediately after fabrication.This indicated that the calcium phosphate phase transition of the amounts of α-TCP used in this study did not critically affect the encapsulated cells either initially or during the entire reaction and culture period.

Effect of cell-laden hydrogels on osteogenic differentiation behavior of cells
After three weeks of culturing the CDHA/GelMA hydrogels laden with MC3T3-E1 cells in osteogenic differentiation medium, the effects of the formed hydrogels on the osteogenic differentiation behavior of the encapsulated cells were examined.ALP activity, calcium ion levels, and expression of osteogenic differentiation marker genes were analyzed, and the results are shown in figure 6.From week 2, cells in the 0.25 and 0.5 wt% α-TCP hydrogel groups had relatively higher ALP activity compared to the other groups.After 3 weeks of culture, the cells in the 0.25 wt% α-TCP hydrogel had the highest ALP activity at 0.49 ± 0.05 mM, as shown in figure 6(a).This is significantly different from the amounts measured in the cells of the other groups (0 wt% α-TCP = 1.55 ± 0.11 µg ml −1 , 0.125 wt% α-TCP = 1.38 ± 0.11 µg ml −1 , 0.25 wt% α-TCP = 0.81 ± 0.07 µg ml −1 , and 0.5 wt% α-TCP = 0.55 ± 0.06 µg ml −1 ).
In terms of calcium ion accumulation, which is indicative of cell mineralization, cells in the 0.25 wt% α-TCP hydrogel had the highest quantity of calcium ions (0.28 ± 0.03 µg) throughout the entire culture period, as shown in figure 6(b).The amount of calcium ions was also measured in a series of negative control samples containing the same amounts of α-TCP used to generate the cell-laden hydrogels (0, 0.125, 0.25, 0.5 wt% α-TCP) but no cells.It was found that the 0.25 wt% α-TCP loaded hydrogel contained less CDHA after calcium phosphate phase transition than the 0.5 wt% α-TCP loaded hydrogel, but more calcium ions than the 0.5 wt% α-TCP loaded hydrogel.
The activation of osteogenesis in the encapsulated cells was further investigated by gene expression analysis.qPCR was used to assess the relative expression of the genes encoding RUNX2, ALP, OPN, and OCN in cells after 3 weeks of culturing, and the results are shown in figures 6(c)-(f).Significantly, for every evaluated marker gene, exposure to the 0.25 wt% α-TCP hydrogel resulted in the highest expression levels.Compared to an α-TCP-free GelMA hydrogel, the 0.25 wt% α-TCP hydrogel induced the following fold increases in expression: 4.1-fold for RUNX2, 12.9fold for ALP, 7.3-fold for OPN, and 6.1-fold for OCN.
All the cell activity analytical results indicated that the 0.25 wt% α-TCP hydrogel has the highest potential for inducing bone regeneration activity in encapsulated cells after calcium phosphate phase transition.

Fabrication of α-TCP/GelMA scaffolds
After the effects of the hydrogels on the osteogenic differentiation behavior of the encapsulated cells were investigated, the optimal quantity of α-TCP to be included in the GelMA hydrogel formulations for cell differentiation was determined to be 0.25 wt%.The next step was perform a printability test to determine the applicability of the hydrogel as a bio-ink.Two types of hydrogels, namely, pure GelMA hydrogel, and 0.25 wt% α-TCP-loaded GelMA hydrogel, were selected for printing 3D structural scaffolds using a customized extrusion 3D printer.
To investigate the printable ranges of the processing parameters (pneumatic pressure, feed rate, and nozzle size) for the fabrication of high-fidelity 3D structural scaffolds, a single line test was performed with cell-free 0.25 wt% α-TCP-loaded GelMA hydrogel.Figures 7(a) and (b) shows the results associated with determining the optimal ranges of pneumatic pressure for producing stable struts per nozzle size.Different ranges of pressures (80-120 kPa) were required to fabricate a stable strut for each nozzle size.In figure 7(a) the range of required pneumatic pressures is shown in figure 7(a), and the strut widths, which were found to depend on the pressure for each nozzle size, are shown in figure 7(b).Each nozzle size produced a different strut width, and the desired strut diameter could be achieved by varying the pneumatic pressure.The 23 G nozzle size was selected due to the possibility of producing struts 418 ± 23 µm wide at 100 kPa of pressure.The feed rate and moving speed were evaluated and the results are shown in figures 7(c) and (d).In brief, a feed rate exceeding 300 mm min −1 could not produce stable struts when using a 23 G nozzle.Finally, the optimal conditions for scaffold fabrication were determined, and the scaffolds were fabricated under a pneumatic pressure of 90-100 kPa and a feed rate of 200 mm min −1 .Two scaffolds were produced using 0 and 0.25 wt% α-TCP-loaded cell-loaded GelMA hydrogels, respectively.
After the optimal printable conditions were determined, the stability and viability of the encapsulated cells during printing were investigated using a live/dead cell assay, and the results are shown in figures 7(e) and (f).There was an abundance of cells emitting green fluorescence, indicating that live cells were present in both types of scaffolds.This confirmed that there was no critical damage to the encapsulated cells during the scaffold fabrication process under the set printing conditions.

Evaluation of cellular activities of the cell-laden α-TCP/GelMA scaffolds
A live-dead cell assay was used to evaluate the viability of the cells encapsulated in the scaffolds after 3 weeks of culture.As shown in the representative images in figures 8(a) and (b), relatively few cells exhibited red fluorescence, which is indicative of dead cells, and the majority of cells showed green fluorescence, which is indicative of live cells in both scaffolds.Surprisingly, cells encapsulated in both scaffolds (with and without α-TCP) exhibited cell growth.Therefore, it can be concluded that the α-TCP-based calcium phosphate phase transition used to produce the α-TCP-loaded GelMA hydrogel did not affect cell proliferation within the 3D-printed scaffold.
Four groups of scaffolds were printed to evaluate osteogenic cellular activities, divided into noncell laden and cell encapsulated scaffolds.All scaffold samples were stained with Alizarin Red S after three weeks of culture, and the results are shown in figures 8(d)-(g).This stain was used to visualize calcium deposits, which appeared as dark red structures and indicate cell mineralization.Both the GelMAbased scaffold and the 0.25 wt% α-TCP-loaded GelMA-based scaffold stained red, indicating that calcium deposits were present.However, the α-TCPloaded GelMA-based scaffold stained more strongly than the GelMA-based scaffold did.Therefore, it was concluded that the 0.25 wt% α-TCP-loaded GelMAbased scaffold contained more cell mineralization than the GelMA-based scaffold.
The mechanical properties of the scaffolds that had been cultured for 3 weeks were investigated by measuring the compressive modulus.As shown in figure 8(c), the compressive modulus of the scaffold containing α-TCP differed considerably from that of the scaffold without α-TCP after three weeks.This indicates that the mechanical properties that were enhanced immediately after the calcium phosphate phase transition were not affected by the longterm culturing.To determine the effect of cell mineralization on the compressive modulus, cell-laden and cell-free α-TCP-loaded scaffolds were studied.In brief, there was no significant difference between the compressive moduli of the two scaffolds.Therefore, although mineralization occurred during the 3 week culture period, it did not significantly affect the mechanical properties of the scaffold.
It was also found that throughout the culture period, the stability of the scaffolds (i.e. its ability to retain its shape) differed considerably.After three weeks of culturing, the α-TCP-loaded GelMA-based scaffold retained its shape; however, the GelMAbased scaffold showed disconnection between the struts, and some of the struts fell apart, resulting in disconnection.Therefore, we concluded that the mechanical properties were affected by the α-TCP calcium phosphate phase transition process in the GelMA hydrogel, because at the 3 week time point, the α-TCP-loaded scaffold exhibited a compressive modulus that was approximately six times higher than that of the scaffold without α-TCP, which affected the ability of the scaffold to retain its shape during long-term culturing.

Discussion
In this study, we aimed to improve the hydrogel properties under conditions that are not harmful to encapsulated cells by using α-TCP phase transition in GelMA hydrogels to generate a bioink that could be used for hard tissue regeneration.Thus, we focused on optimizing the mechanical properties of the final hydrogel, as well as the viability and osteogenic differentiation behavior of the encapsulated cells.Thus, the reaction components and conditions were optimized, as shown in figure 1.As mentioned above, α-TCP was added to the hydrogel to improve its mechanical properties.More specifically, α-TCP undergoes phase transformation via hydrolysis in neutral aqueous solutions at 37 • C, which is known as a calcium phosphate phase transition [37].During this phase transformation, α-TCP transforms into crystalline CDHA, which becomes entangled more easily and may confer improved mechanical properties [37,52].
Therefore, we aimed to determine whether the inclusion of α-TCP in the GelMA hydrogel results in a typical calcium phosphate phase transition, regardless of the amount loaded, over a set period.For this, we fabricated GelMA hydrogel discs that contained different amounts of α-TCP (0, 0.125, 0.25, 0.5 wt%) and immersed them in cell culture medium for 36 h at 37 • C. XRD analysis showed that after 36 h, regardless of the amount of α-TCP in the GelMA hydrogel, CDHA was present, and no α-TCP was detected.In addition, the transition of α-TCP to crystalline CDHA was confirmed by SEM (figure 2).It was clear that at least 36 h were needed to complete the calcium phosphate phase transition of α-TCP in the hydrogel, and it was confirmed that GelMA did not disturb the reaction.
Swelling is an important factor in the ability of hydrogels to retain their structure over the long term [15,17,53].Thus, the effect of the calcium phosphate phase transition on hydrogel swelling was investigated.As shown in figure 3, hydrogel swelling decreased in a manner proportional to the increase in the amount of α-TCP (fully transformed to CDHA) in the GelMA after the calcium phosphate phase transition.It is known that the water retention ability of CDHA is better than that of α-TCP because the ability to absorb water increases linearly with the surface area of apatite [54,55].Thus, we confirmed that the presence of CDHA could reduce hydrogel swelling by absorbing water.
To verify the effects of the calcium phosphate phase transition on the mechanical properties of the GelMA hydrogel, its compressive modulus was measured.The results are shown in figure 4, it is clear that, during calcium phosphate phase transition, the mechanical properties of the hydrogel increased as more α-TCP was loaded.When lower amounts of α-TCP were loaded, it could not act as a filler to improve the mechanical properties by loading only.However, as shown above, after the calcium phosphate phase transition, the compressive modulus increased significantly by approximately eight times for the sample containing 0.5 wt% α-TCP.Even the 0.25 wt% α-TCP group showed an increase in compressive modulus, but only after the calcium phosphate phase transition.Therefore, we concluded that α-TCP could sufficiently reinforce the mechanical properties of hydrogels when added in all tested quantities.We assumed that the observed dramatic changes in the mechanical properties of the hydrogels were due to the crystalline phase of the bioceramic, which could increase the interactions between the bioceramic and hydrogel through the larger surface area.During the calcium phosphate phase transition, the dissolution of α-TCP results in an increase in calcium and phosphate ions, which precipitate to form needle-shaped CDHA, which has a greater surface area than α-TCP [36][37][38].
Developing a with advantageous mechanical properties and supporting cell viability was an essential part of this study.The effects on cells during and after the calcium phosphate phase transition were investigated once the improved hydrogel properties were verified.According to the observations in figure 5(a), during the calcium phosphate phase transition, the DNA quantities of the group that had been loaded with α-TCP had decreased without regard to the amount of loaded α-TCP.Even in figure 5(b), the α-TCP loaded group showed lower DNA amounts throughout the whole cultivation time than the pure GelMA group.This demonstrated that the amount of DNA present in the cells during the phase transition and during cultivation was significantly affected by the calcium phosphate phase transition (from α-TCP to CDHA).
For the reason for this decrease, we assumed some of the respectable reasons.First, one of the factors can, in the initial stages, be the calcium ion generated during the phase change of calcium phosphate.As shown in figure 6(b), even if the calcium ion produced might not exhibit a significant level of concentration, it could still have an impact on the DNA amounts of cells that have been encapsulated.Since it is already known that calcium accelerates cell death via apoptosis in the mitochondria by interacting with normal metabolism, a rising calcium ion concentration could lead to calcium ion influx into the mitochondria of the encapsulated cells [50,51].However, we could not be convinced fully, yet, because even figure 6(b) demonstrated that the calcium ion generated during cultivation failed to reach a sufficiently high concentration to directly influence cells and induce death.Secondly, we adopted the assumption that DNA binding and the calcium ion were the other possible causes [56,57].The reduction in DNA amounts could also be affected by the interaction between calcium ions and DNA, which proposed that the hydrogel's calcium ions and calcium-bonded DNA led the DNA quantities to be lower than they were supposed to be.
However, as we just indicated, all of them are only hypotheses that we could think of as possible causes of DNA decrease and that require further investigation to be fully verified.To fully comprehend the relationship between the calcium phosphate phase transition products and cell viability, it is necessary to investigate the signaling pathways that cells use to die to identify if they are necrotic or apoptotic.Therefore, we will continue studying how this phase transition affects cell viability.
To further examine the cell viability, the DNA of the encapsulated cells was quantified.After one week of culturing in an osteogenic differentiation medium, cells encapsulated in each type of hydrogel (i.e. in the presence of different amounts of α-TCP) showed an increase in the quantity of DNA compared to that at 0 h, as shown in figure 5(b).We postulated that after 36 h of the calcium phosphate phase transition, the ions generated by the reaction may reduce cell viability.However, after 36 h, the live encapsulated cells in each hydrogel proliferated and differentiated normally.It is also possible that the cellular activity was due to the GelMA matrix, which has outstanding biological activities [58].
However, during the following two and three weeks of culture, the DNA level of every group, regardless of the amount of α-TCP, gradually decreased.This phenomenon occurred in every group, including those without α-TCP; thus, it was unlikely to be caused by α-TCP.While decreases in cell proliferation were observed at 2-3 weeks, there were significant increases in osteogenic differentiation in the cells of the α-TCP-loaded groups compared to the cells of the GelMA group (without α-TCP), in every analysis performed (figure 6).Notably, cells with 0.25 wt% α-TCP group showed osteogenic differentiation activity throughout the entire period.Based on the above findings, we concluded that the 0.5 wt% α-TCP-loaded hydrogel has more advanced mechanical and swelling properties, but that the encapsulated cells in this hydrogel cannot survive due to the high concentration of calcium ions generated during the calcium phosphate phase transition.The 0.25 wt% α-TCP-loaded hydrogel also negatively affected cell viability during the calcium phosphate phase transition.However, after the calcium phosphate phase transition, cell proliferation occurred, and even though the DNA level decreased during the culture period, it recovered and reached a level similar to the initial level.Also, the 0.25 wt% α-TCP-loaded GelMA hydrogel had an improved compressive modulus and reduced swelling, indicating that it retains its structural fidelity.
Therefore, we selected a 0.25 wt% α-TCP-loaded GelMA hydrogel as the candidate bioink and proceeded to optimize the printing conditions.We determined the optimal printing conditions for the three parameters: nozzle size, pneumatic pressure, and feed rate.Cell-laden scaffolds were printed from the hydrogel under optimized conditions, and the cell viability of the encapsulated cells was investigated using a live/dead cell assay for toxicity or any damage that could result from fabrication.As shown in figure 7, there was no critical damage to the cells during printing.Moreover, it is clear that including 0.25 wt% α-TCP in the GelMA hydrogel improved the fidelity of the scaffold structure.
After 3 weeks of culturing in osteogenic differentiation medium, cell growth was observed in both GelMA and 0.25 wt% α-TCP-loaded GelMA hydrogel scaffold samples (figure 8).This cell growth is likely supported by the GelMA hydrogel.It was observed that cell growth in the 0.25 wt% α-TCP-loaded GelMA hydrogel scaffold samples was reduced during the culture period compared to that in the GelMA J Kim et al hydrogel scaffold samples, and this was likely due to the calcium ions generated by the calcium phosphase transition.However, after three weeks, definite cell growth in the 0.25 wt% α-TCP-loaded GelMA hydrogel scaffold samples.Moreover, more cell mineralization, which was stained darker red with Alizarin Red S, occurred in the α-TCP-loaded GelMA hydrogel scaffold samples than in the samples without α-TCP.In figures 8(d)-(g), the 0.25 wt% α-TCPloaded GelMA hydrogel scaffold, regardless including cells or not, was stained in red color because of the contained α-TCP in the hydrogel.However, notably, after 3 weeks of cultivation in osteogenic differentiation medium, the cell-laden 0.25 wt% α-TCP loaded GelMA scaffold showed darker red staining compare to the scaffold without cells.Therefore, we could conclude that additives of α-TCP, which phase transferred to CDHA during cultivation, helped to improve cell mineralization.
The compressive modulus of the cell-laden scaffolds was investigated to determine the effects of cell mineralization on compression and to investigate the shape retention of the scaffolds after immersion in an aqueous solution for a long period (3 weeks).As shown in figure 8(c), cell mineralization did not affect the mechanical properties of the cellladen scaffolds; however, there were significant differences in the mechanical properties between the α-TCP-loaded and non-loaded GelMA hydrogel scaffolds.The 0.25 wt% α-TCP-loaded group showed a six-time stronger compressive modulus, even after 3 weeks in cell medium, and because of this property, the scaffold preserved its original shape without any fragmentation of struts.The non-loaded GelMA hydrogel scaffold could not retain its shape, and the struts on the top were disconnected, causing the center of the scaffolds to collapse.
In this study, we determined the optimal amount of α-TCP to be included in hydrogel formulations to improve the mechanical properties of hydrogels and maintain a non-harmful cellular environment.The findings of this study demonstrated that the GelMA hydrogel with 0.25 wt% α-TCP has beneficial properties and potential as an effective bioink for hard tissue regeneration.

Conclusions
In this study, α-TCP was added to and mixed homogenously in the GelMA hydrogel to overcome the limitations of the current hydrogels.The calcium phosphate phase transition that transformed α-TCP into CDHA was found to enhance the mechanical properties of the GelMA hydrogel up to eight times.Cell viability and activity assays were performed to determine the optimal amount of α-TCP to be included in hydrogels that conferred the desired mechanical properties without causing critical damage to encapsulated cells.The α-TCP-loaded hydrogel exhibited the potential to induce osteogenic differentiation of encapsulated cells.The optimal printing conditions for α-TCP-loaded GelMA hydrogel composites were determined.There were significant differences in the mechanical properties and structural stability between the α-TCP-loaded GelMA and the non-loaded GelMA hydrogel-printed scaffolds.Therefore, we anticipate that the α-TCP-loaded GelMA-based bioink developed in this study can be utilized to induce osteogenesis during hard tissue regeneration.

Figure 1 .
Figure 1.Schematic image of the osteogenesis-enhancing α-TCP/GelMA bioink and the CDHA/GelMA produced via the calcium phosphate phase transition process.

Figure 2 .
Figure 2. Phase transition during the calcium phosphate phase transition process of encapsulated α-TCP to CDHA in the GelMA solution.X-ray diffraction patterns of the sample hydrogels at 0 h (a) and at 36 h (b), and FE-SEM images of the sample hydrogels at 0 h (c) and at 36 h (d) are shown.Blue arrows indicated the added α-TCP, before the phase transition occured.Red arrows indicated the formed CDHA, after the phase transition.

Figure 3 .
Figure 3. Water content of the hydrogel samples according to the amount of α-TCP at (a) 0 h and (b) 36 h ( * p < 0.05; n.s.= not significant).

Figure 5 .
Figure 5. Cell viability of encapsulated cells in each hydrogel: (a) CCK-8 analysis of encapsulated cells during calcium phosphate phase transition, and (b) DNA quantification for entire culturing duration.

Figure 8 .
Figure 8. Live/dead assay results for the fabricated scaffolds after 3 weeks of culturing: (a) GelMA and 0.25 wt% α-TCP/GelMA.A comparison of the compressive modulus values of the scaffolds after the 3 week culture period is shown in (c).In vitro osteogenic cellular activities were assessed using Alizarin Red S staining: without cells (d) GelMA (e) 0.25 wt% α-TCP/GelMA and cell encapsulated (f) GelMA and (g) 0.25 wt% α-TCP/GelMA.