Improving pore size of electrospun gelatin scaffolds containing graphene oxide using PEG as a sacrificial agent for bone tissue engineering

Abstract In this study, polyethylene glycol (PEG), as a sacrificial agent, and gelatin containing graphene oxide (GO) are co-electrospun to prepare scaffolds. Based on pore size measurements, removal of PEG leads to an increase in the pore size of modified scaffolds, resulting in easier infiltration. Moreover, noticeable elastic modulus of GO increased the mechanical strength. Alizarin red staining indicated proliferation and differentiation of human bone marrow-derived mesenchymal stem cells seeded on the modified scaffolds. Thus, gelatin scaffold containing 0.5% GO with 20% PEG at the ratio of 80:20 for gelatin–PEG was selected as the optimal scaffold for bone tissue engineering approaches. Graphical Abstract


Introduction
Bone defects are the main musculoskeletal disorders affecting the quality of life.In case of critical bone loss, selfhealing does not occur, particularly in elder patients [1,2] .These defects may occur due to surgery, tumors, chronic infectious diseases, and acute injuries causing open fractures [3,4] .Tissue transplantations such as autografts and allografts have limited uses due to the additional surgical operation, morbidity of the donor site [5] , infection risks, and disease transmission [6] .In recent decades, bone tissue engineering has been introduced to imitate bone formation procedures by harnessing stem cells, scaffolds, and biological factors [7] .
The scaffold should mimic the natural extracellular matrix (ECM), support cell function, such as cell attachment, migration, and differentiation, and facilitate the transport of regulatory factors and nutrients [8,9] .Several conventional techniques have been employed to fabricate scaffolds, including freeze-drying, solvent casting, particulate leaching, thermal-induced phase separation, and electrospinning [10] .
Electrospinning is a versatile method for creating micro and nanofibrous scaffolds.Electrospun scaffolds have been widely utilized regarding their facile functionalization, availability, cost-effectiveness, and porosity.Also, they provide a large surface area-to-volume ratio, which leads to favorable mechanical properties and an increase in cell adhesion [11,12] .However, the small pore size of the electrospun scaffolds affects the seeded cells' deep proliferation, limits nutrients and waste transfer, and confines vascularization and tissue integration of the scaffolds [13] .In this respect, some methods, such as salt leaching and cryogenic electrospinning [14] , sacrificial fibers [15] , ultrasonication [16] , and gas foaming [17] can support cell function by improving pore size.
Using a sacrificial agent during the electrospinning process and its removal selectively is a prevalent and efficient strategy to increase the pore size in the electrospun scaffolds without distorting the structure [18] .
Aghajanpoor et al. studied the effect of simultaneous electrospinning of polycaprolactone (PCL) and polyethylene oxide (PEO), and the results showed that using PEO as the sacrificial fibers led to larger pores [19] .Moreover, He et al. found that PEO removal can promote the porosity and cellular infiltration of fish collagen-PCL nanofibers [20] .In another research, Skotak et al. investigated gelatin-polyethylene glycol (PEG) (with 1% PEO for its spinning ability) electrospun scaffolds.They reported that the pore size of the scaffolds was increased by PEG removal, which facilitates the penetration and growth of fibroblast cells [21] .
Biomaterials are also considered a crucial factor in preparing scaffolds.Natural polymers such as collagen, gelatin, elastin, fibrinogen, and laminin can mimic the ECM by creating biodegradable scaffolds with low inflammation risks and promoting cell adhesion [22][23][24] .Many scientists have focused on gelatin due to its availability, biocompatibility, and biodegradability.Furthermore, gelatin has a biological origin, and its cost is relatively lower than collagen [25] .However, some restrictions, such as fast and uncontrollable degradation and weak mechanical strength, should be noticed for its applications in bone tissue engineering [26] .For instance, Jaiswal et al. indicated that hydroxyapatite nanoparticles on the surface of gelatin-PCL scaffolds improve their mechanical properties [27] .In another study, Jalaja et al. reported that gelatin nanofibers containing graphene oxide (GO) have more tensile strength than gelatin nanofibers [28] .Likewise, GO regulates the Wnt/b-catenin pathway, which plays a critical role in the regeneration of bone tissue [29] .Therefore, improving the mechanical strength and inducing differentiation may make GO as an advantageous candidate for bone tissue engineering purposes.
In this study, the effect of PEG as a sacrificial agent on the pore size improvement of gelatin electrospun scaffolds is investigated.Also, due to its remarkable elastic modulus [30] , GO was added to modify gelatin scaffolds' mechanical properties.Furthermore, cell studies were performed to confirm the biocompatibility, lack of toxicity, and differentiation of mesenchymal stem cells into the bone cells on the scaffolds.

Scaffolds preparation
The solution of gelatin (25 w/v%) was prepared in a 40:60 mixture of acetic acid/water [31] , and the mixture was stirred using a magnetic stirrer at 30 C for 12 h.PEG was dissolved in ethanol 90% by stirring at room temperature for 2 h to prepare 20 and 30 w/v% solutions; the concentrations were based on our pretests (data not shown).The polymer solutions (gelatin and PEG) were transferred into the separated 10 mL syringes and simultaneously electrospun (Figure 1) at the following conditions: solution flow rate ¼ 0.4 mL/h, applied voltage ¼ 20 kV, tip-to-drum distance ¼ 12 cm, syringes distance from each other ¼ 5 cm, and drum speed ¼ 600 rpm [19,32] .These conditions were the same for all the electrospun scaffolds.
As a modification, 20 w/v% PEG solution was blended with 25 w/v% gelatin at the ratio of 80:20 Gel-PEG under magnetic stirring at 30 C for 2 h to obtain a homogeneous Gel-PEG solution and then transferred into one syringe.The same procedure was followed for 30 w/v% PEG solution also to optimize PEG concentration.Moreover, the scaffolds with the ratio of 70:30 Gel-PEG and both PEG concentrations were analyzed to determine the optimal pore size of the scaffold.These two ratios were selected according to our pretests (data not shown).
In the next step, GO solution (10 mg/mL) was added to Gel-PEG solutions at different PEG concentrations and Gel-PEG ratios by continuous stirring at 30 C for 1 h [33] to prepare the suspensions with 0.5 and 1 w/v% of GO for Gel-PEG-GO electrospun scaffolds preparation.In addition, for better distribution of GO in the suspension, before electrospinning, the suspension was homogenized in an ultrasonic water bath at a power of 90% for 20 min.Gel-PEG-GO suspension was then electrospun through a 10-mL syringe.In this study, an electrospun gelatin scaffold was prepared at a concentration of 25 w/v% as the control group.
The gelatin nanofibers were crosslinked by exposing them to 1 v/v% GTA vapor in a desiccator for 48 h.Then, the samples were placed in a 5-mL beaker filled with t-BuOH and kept in a water bath at 60 C to remove PEG from the scaffolds.After 24 h, the hot solution was removed immediately, and the samples were lyophilized for 24 h to remove the residual solvent [21] .

Characterization of the electrospun scaffolds
The morphology and physical structure of the electrospun scaffolds were investigated using a scanning electron microscope (SEM; MIRA3, TESCAN) operating at an accelerating voltage of 15 kV.SEM images were taken from different parts of each scaffold, and the electrospun fiber diameters and pore size distribution were estimated using ImageJ software [19] (further details of this method is provided in the Supplementary Material).The chemical structure of the electrospun scaffolds and the removal of PEG were investigated using Fourier Transform Infrared spectrometer (FTIR; Perkin-Elmer Spectrum).

Mechanical strength
Some of the electrospun scaffolds were selected for tensile assay based on pore size optimization.They were cut into 60 Â 10 mm 2 pieces and placed between two clamps of the testing machine (STM20, Santam) at a distance of 2 cm from each other and pulled at a speed of 10 mm/min.Three samples from each group were assessed, and the average amount was reported.

Cell culture analysis
Human bone marrow-derived mesenchymal stem cells (hBMSCs) were obtained from Royan Institute and cultured in cell culture flasks containing DMEM medium with 15% FBS and 1% penicillin/streptomycin, kept in an incubator at 5% CO 2 and 37 C.The culture medium was refreshed every other day.
The prepared scaffolds with 15-mm diameter were placed in 24-well culture plates and sterilized by immersing twice in ethanol 70 v/v% for 15 min.Then, ethanol was removed by rinsing with PBS and DMEM medium, respectively, for 15 min.Afterward, the cells were detached by 0.25% trypsin-EDTA at 80% confluency, and 50,000 cells (passage 3) were seeded on each scaffold.They were incubated at 5% CO 2 and 37 C for 45 min to obtain better cell adherence to the scaffold.Eventually, 650 lL of culture medium was added to the prepared samples and returned to the incubator at 5% CO 2 and 37 C.The culture medium was refreshed every 2 days.In this study, the same numbers of hBMSCs were cultured inside a scaffold-free well as the control group.

MTT assay
Viability and proliferation of the cultured cells on the scaffolds were assessed using MTT assay after 1, 4, and 7 days.Briefly, the culture medium was removed, replaced with a 5:1 mixture of DMEM medium/MTT solution, and incubated at 5% CO 2 and 37 C for 2 h.Then, the medium was removed, formazan crystals were dissolved in dimethyl sulfoxide, and the absorbance at 545 nm was measured using an ELISA reader (State Fax 3200, Awareness Technology).Each group had three replications.

Alizarin red staining
Calcium secretion and mineralization were evaluated by staining the cultured samples using alizarin red after 7 and 14 days.Briefly, the groups of control and scaffold-containing cells (with three replications) were cultured in culture medium containing 15% FBS and osteogenic differentiation medium (10-8 M dexamethasone, 50 mg/mL ascorbic acid, 10 mM glycerol phosphate) containing 10% FBS, respectively.The culture medium was completely changed every other day.The following steps were applied to stain the cells: first, the culture medium of the samples was completely removed.Then, the cells were washed with PBS and stabilized by exposing them to methanol for 15 min.In the next step, after removing the methanol, alizarin red stain solution with a concentration of 1% (pH ¼ 4.2) was added to the wells.The cells and scaffolds were in contact with the solution for 10 min.Then, the solution was drained gently, and the samples were washed slowly with distilled water five times to remove the excess dye.Finally, the samples were observed and photographed using a reverse light microscope (Olympus IX71, Tokyo, Japan).

Statistical analysis
Statistical significance was determined by a p-value<0.05.The provided results have been presented as mean ± standard deviation (SD).

Characterization of the electrospun scaffolds
Morphology characterization of Gel-PEG electrospun scaffolds with polymer solutions in separate syringes was investigated by SEM. Figure 2 shows that the presence of beads between Gel fibers disrupted the formation of homogenous and favorable porous structure because the PEG solution could not be electrospun alone.
The porous structure and interconnected pores of the electrospun scaffolds can increase the surface-to-volume ratio and improve cell adhesion to the scaffolds.These enhancements are because of providing enough space for nutrients and oxygen transition and angiogenesis in the newly formed bone tissue [14] .Hence, to prepare an electrospun scaffold without any beads, PEG solution was blended with Gel solution before transferring into the syringe.The "modified scaffolds" with Gel 25%/PEG 20% (w/v) at 80:20 ratios were chosen as a sample to investigate the possibility of fiber formation.SEM images of the modified scaffolds showed fibers with random orientation, porous structure,  and interconnected pores that are essential for bone tissue engineering purposes (Figure 3).The average fiber diameter for these scaffolds was 0.690 ± 0.154 lm.
The modified scaffolds after crosslinking were studied by obtaining SEM images, which showed the morphology changes (Figure 4).The average fiber diameter, in this case, was 1.022 ± 0.293 lm.
Chemical characterization was performed using FTIR to evaluate the removal of PEG from the modified scaffolds.Figure 5 illustrates FTIR spectra, including Gel, GTA, PEG, and modified scaffolds before and after the removal of PEG.The peaks at 843, 947, 1,280, 1,342, and 2,889 cm À1 are related to PEG [34,35] , which are also observed in the modified scaffold spectra before removal.The absence of these peaks in the modified scaffold spectra after the removal indicates complete washing of PEG from the modified scaffold.Therefore, PEG removal was confirmed by FTIR.

Pore size evaluation
The pore size was evaluated by preparing SEM images of the modified scaffolds before washing (PEG removal) and after lyophilizing (Figure 6).The results showed that the pore size of the modified scaffold was increased after the removal of PEG using t-BuOH solvent.Generally, the presence of PEG during the electrospinning of gelatin and crosslinking may have created a spatial barrier and prevented the formation of several connections.It is worth mentioning that freeze-drying is also a method to have larger pores [36] .However, SEM images of this study illustrated no significant difference between the pore size before and after the lyophilizing process (data not shown).

Pore size optimization
The scaffolds were electrospun in four groups at different PEG concentrations and Gel-PEG ratios, and their SEM images were prepared (Figure 7) to determine the optimal PEG concentration leading to wider pores.The average pore size of all groups was estimated before and after PEG removal, and the measurements showed an increase in the pore size after PEG removal (Table 1).Notably, the pore size of gelatin scaffolds as a control group was measured after freeze-drying, and the results were estimated as 5.500 ± 5.537 lm 2 .Comparing the results indicated that gelatin scaffolds containing the sacrificial agent of PEG with a concentration of 20% possessed larger pores than the concentration of 30% at both Gel-PEG ratios.A higher percentage of PEG prevented proper electrospinning.As mentioned earlier, PEG alone cannot be electrospun, and its high percentage in the solution leads to fibers scattering and improper collection on the drum.In this research, using 20% PEG may lead to suitable crosslinking of the fibers, demonstrates larger spaces between them, and makes it the optimal concentration for the following studies.Thus, PEG at a concentration of 20% and two ratios of 80:20 and 70:30 Gel-PEG were selected as the optimal samples.

Mechanical properties
According to previous studies, GO addition increased the strength and modulus [33] , thereby improved the mechanical properties of gelatin scaffolds.In this study, 0.5 and 1 w/v% GO were added to the scaffolds separately to improve the mechanical properties of gelatin.The tensile test results (Table 2) show that the average Young's modulus of 80:20 Gel-PEG scaffolds was significantly increased in the presence of GO, while it had no effects on the modulus of 70:30 Gel-PEG scaffolds.The fragile structure of 70:30 Gel-PEG ratio had lower mechanical strength, which might be due to the higher amount of removed PEG from the scaffold.Therefore, the addition of GO to 70:30 scaffolds had no significant effect on the improvement of Young's modulus.On the other hand, comparing   (c, d) Group 2; (e, f) Group 3; and (g, h) Group 4 (the groups are according to Table 1).
the stress (Figures 8 and 9) at two ratios of Gel-PEG, containing 1% GO indicates that at 30% strain, the stress is more pronounced at the ratio of 80:20 than 70:30 (2.1 and 1.3 MPa, respectively).This is because less amount of PEG as a sacrificial agent in the gelatin scaffold induces more mechanical strength.
In this respect, Campiglio et al. [37] have confirmed the effect of crosslinking on enhancing the mechanical strength of gelatin scaffolds.Hence, crosslinking via GTA could be another reason for changing the strength and modulus, which might be due to the increasing average fiber diameter of the scaffolds in this study from nanometric to micrometric scale.Therefore, 80:20 Gel-PEG (20%) scaffolds containing 0.5 and 1 w/v% GO were selected as the optimal samples.

MTT assay
MTT assay was performed on three groups containing different amounts of GO (Table 3) and a control group (hBMSCs cultured in a scaffold-free well) to evaluate the cell viability of the selected optimal scaffolds.According to the MTT results (Figure 10), after 1, 4, and 7 days, the viability percentage of the live cells on the scaffolds containing GO was more than the control group.The explanation is that the presence of GO affects the improvement of cell adhesion and proliferation [38] .Also, higher amounts of GO (>2%) induce cytotoxicity [39] .Cell viability in 80:20 Gel-PEG (20%) scaffolds containing 0.5% GO, was slightly more than the scaffolds with 1% GO, confirming the cytotoxicity of GO.In addition, the cytotoxic effect of 1% GO resulted in a nonsignificant difference between the responses of 80:20 Gel-PEG (20%) scaffolds containing 1% GO and gelatin scaffolds.On the other hand, a significant difference between the percentage of live cells in the 80:20 Gel-PEG (20%) scaffolds containing 0.5% GO and gelatin scaffolds confirmed the positive impact of GO on increasing the adhesion, penetration, and proliferation of hBMSCs.Therefore, 0.5 w/v% GO was the optimized condition for cell culturing.Besides, this assay revealed that the presence    2).
of pores in the scaffolds has contributed to higher cell adhesion.Moreover, cell proliferation in the pores increased compared to the control group due to the possibility of nutrient transition in the porous 3D structure of the scaffolds.Therefore, in addition to viability, the cells could penetrate the structure and proliferate.These processes make the scaffold a suitable bed for biological interactions.

Alizarin red staining
Alizarin red staining was investigated to analyze the production of calcium by the cells and its presence in the scaffolds (Figure 11).According to the control group images, alizarin red staining confirms osteogenesis induction in hBMSCs.After 14 days, more red crystals were visible in the control group, indicating the proper function of the osteogenic medium and induction of bone differentiation.Comparing all the groups on day 7 revealed that hBMSCs in the scaffolds containing GO were more differentiated into bone cells.As mentioned earlier, in addition to increasing the mechanical strength of the scaffolds, GO induced and improved bone differentiation of hBMSCs [40] .The presence of more GO particles in 80:20 Gel-PEG (20%) scaffolds containing 1% GO caused more cells to differentiate into osteoblasts.However, on day 14, the toxicity effect of GO could be observed for this group, which is more compared with   3).The control group is hBMSCs cultured in a scaffold-free well.
Figure 11.Alizarin red staining images (the groups are according to Table 3).The control group is hBMSCs cultured in a scaffold-free well.
the scaffolds containing 0.5% GO.Eventually, the gelatin scaffold containing lower amounts of GO particles resulted in more differentiation and less cell death, making it the optimal scaffold in this study.

Conclusion
We investigated the enhancement of pore size and mechanical properties of gelatin scaffolds.PEG was used as the sacrificial agent to increase the pore size, and the porous Gel-PEG scaffold was successfully prepared with simultaneous electrospinning.The results of pore size optimization showed that optimal PEG concentration led to larger pores.Also, GO was added to the scaffold to improve mechanical strength.Based on the obtained results, GO could increase Young's modulus of the scaffolds in an optimized ratio.Moreover, cell studies indicated biocompatibility and differentiation of hBMSCs for scaffolds containing GO.Hence, the scaffold introduced here provides new approaches for improving gelatin scaffolds and their applications in bone tissue engineering.

Figure 2 .
Figure 2. SEM image of Gel-PEG electrospun scaffold in two syringes containing gelatin solution and PEG solution separately.

Figure 3 .
Figure 3. SEM image of the modified scaffold.

Figure 4 .
Figure 4. SEM image of the modified scaffold after crosslinking.

Figure 6 .
Figure 6.SEM images of the modified scaffold: (a) before and (b) after PEG removal.

Figure 10 .
Figure 10.MTT assay results (the groups are according to Table3).The control group is hBMSCs cultured in a scaffold-free well.

Table 1 .
The pore size of different scaffolds, calculated using Image J software.

Table 2 .
Tensile test results of different scaffolds.

Table 3 .
Different groups for MTT and osteogenesis assay.