At the first, HNTs (4 mg/ml; Sigma Co.) were added to ZA supersaturated solution (4 mg/ml; Sigma Co.) mixed in deionized water under vacuum situation. After 20 min, the suspension was removed, stirred vigorously overnight and then centrifuged for 20 min with 6,000 rpm. The obtained precipitates were washed thoroughly and then freeze dried.

Drug release behavior of the scaffolds were measured by immersing and incubating the scaffolds in PBS at 37°C in a shaker incubator (90 rpm). The scaffold-released ZA level was examined by UV-Visible spectroscopy (NANODROP 2000c; Thermo Scientific Company, United States) at the wavelength of 210 nm at many times point (24, 48, 72 h and 7, 14 and 21 days).

All tested were performed in triplicates and the data were presented as the mean ± standard deviation (SD). One-way ANOVA was used to evaluate the differences between the samples. Statistically significant levels were considered to be p < 0.05.

The morphology and pore size of the prepared scaffolds are shown in Figure 1. As expected, an interconnected network structure was observed in the scaffolds that is in line with the our previous study in which a gelatin-based scaffold containing HNTs and the hydrophilic drug strontium ranelate was synthesized, and the results showed a uniform structure with well-sized and interconnected cavities (Abdollahi Boraei et al., 2020b). Since the pore size distribution doesn’t significantly change with the addition of ZA, the prepared scaffolds show almost similar surface morphologies which consisted of homogenous porosities.

The pore size results showed that the Gel scaffold has the minimum pore size with the mean value of 80.94 ± 43.80 µm, while the pore size elevated to 175.57 ± 21.09 µm in Gel-HNT and 257.89 ± 16.11 µm in Gel-HNT/ZA scaffolds. As expected, the augmentation of HNTs and HNT/ZA leads to create bigger pores, which is in a desire range for BTE applications. In the studies, different numbers have been reported for the appropriateness of the pore size of scaffolds in tissue engineering, for example, in the article of Li et al., The number of 300–400 microns has been selected as the appropriate range (Li et al., 2016; Rashedi et al., 2021). Lee et al. Also considered the number 500 microns to be suitable for adhesion, differentiation, and proliferation of cells inside the scaffold (Lee D. J. et al., 2019). In our studies reported that the scaffolds with larger pore size (>100 µm) provide better matrices for bone regeneration (Abdollahi Boraei et al., 2021b). In general, this number (porosity) should be suitable for delivering nutrients into the scaffold and removing waste from the scaffold (Tolba et al., 2010; Ghazalian et al., 2022).

Table 1 displays the porosity percentage of the prepared scaffolds. A similar upward trend is observed for porosity. Increasing pore size and porosity probably can effects on the electrostatic repulsion forces between carboxyl and hydroxyl groups in gelatin (Type A; pI = 7–9) and HNTs, since both HNTs and gelatin are negatively charged at pH 7 (Abdollahi Boraei et al., 2021a).

The equilibrium water uptake results of the scaffolds are brought in Figure 2. The water uptake for the Gel-HNT/ZA scaffold was the highest (5.02 ± 0.316 (g/g)), while this value for the Gel and Gel-HNT scaffolds decreased to 4.66 ± 0.33 and 1.67 ± 0.369 (g/g), respectively. This increase can be related to the larger pore nature in the Gel-HNT/ZA scaffold and proved by such studies that the microstructure can affect the swelling ratio (Mirahmadi et al., 2013). The more swelling ratio resulted in more water adsorption that is suitable for dipper diffusion of nutrition and better removing of waste from the matrix (Mirahmadi et al., 2013).

FTIR analysis examines the chemical interaction of scaffold components that can determine the three-dimensional structure, the drug release properties, and the biological properties of the scaffold. Figure 3 shows the FTIR graphs of the scaffolds. As expected, in the gelatinous scaffold graph, characteristic gelatin peaks, including amide I, II, III, amide B, and amide A peaks, were observed at 1,698, 1,543, 1,236, 3,067, and 3,421 cm⁻¹, respectively. In the previous study, where the gelatin-based scaffold was made by the freeze-drying method, the peaks mentioned above were observed in the scaffold structure and examined (Abdollahi Boraei et al., 2020a). After the HNTs are added to the scaffold, most of its peaks are overlapped with gelatin, a phenomenon that has been observed in the previous study (Abdollahi Boraei et al., 2021b). Only peaks of gelatin amide A (3,421 cm⁻¹) and O-H stretching of HNTs (3,695 cm⁻¹) shifted slightly, indicating successful incorporation of HNTs into the gelatin structure and even reinforcing the possibility of hydrogen bonding interactions between them. With the addition of ZA to the structure and synthesis of Gel-HNT/ZA nanocomposite scaffolds, new peaks appeared, these peaks include the following: P-O bond (1,300 and 1,322 cm⁻¹), free hydroxyl group (3,492 cm⁻¹), hydroxyl group (2,500–3,300 cm⁻¹) (Paris et al., 2015; Chen et al., 2020) that is the reason for the successful addition of ZA to the structure.

Figure 4 shows the diffraction pattern of a gelatinous scaffold with a wide peak at 20.25^°, which is related to the crystalline structure of the gelatin triple helical, which has been reported in several studies (Jalaja et al., 2015). In general, the HNT pattern has many sharp peaks (Liao et al., 2016). A peak at 20.3^° is observed which is one of the characteristic peaks of HNT that related to (020/110) plane (Liao et al., 2016). The location of this peak has not changed due to the scaffolding production process, which indicates that the HNT layers do not change and that the HNT and gelatin do not intercalate (White et al., 2012). As a result of the addition of ZA to the Gel-HNT scaffold, a small increase in the intensity of the Gel-HNT scaffold peaks occurred, which could be due to the penetration of ZA into the silicate layers of the HNTs. However, due to the absence of ZA characteristic peaks in the graph, it can be concluded that ZA is not completely crystalline in structure. This may be due to the very low concentration of ZA (˃10^–4 mol/L) in the structure that is in line with the results of a previous study in which ZA was loaded into a composite structure and coated on a magnesium implant containing strontium (Li et al., 2019).

Table 2 shows the results of the mechanical compressive strength test of scaffolds. As can be seen, the Gel scaffold has the lowest mechanical strength (σ = 10.27 ± 1.58 MPa), followed by the Gel-HNT scaffold with (σ = 20.92 ± 2.16 MPa) and Gel-HNT/ZA scaffold with (σ = 26.18 ± 2.9 MPa). Despite the increase in pore size in Gel-HNT and Gel-HNT/ZA scaffolds, which should have led to a decrease in mechanical strength, the strength of these samples has increased. The reason for this can be related to the presence of HNTs, which is a ceramic with very high mechanical properties, and the presence of HNT in the structure may lead to the bearing and transfer of force throughout the structure and increase the mechanical properties of scaffolds containing HNTs. In several studies, the effect of adding HNT to biopolymers was investigated, all of which showed an increase in the mechanical properties of the resulting structure (Liu et al., 2013; Liu et al., 2015). Increasing the strength of synthesized tissue engineering scaffolds is considered as one of the practical advantages because scaffolds must act as temporary support against stresses and forces until the tissue is completely repaired (Liu et al., 2015).

Recent studies have shown that local delivery of osteogenesis drugs with a controlled behavior can be used as a way to improve large and severe bone defects with low side effects (Boraei et al., 2020). The release behavior of ZA was initiated by a low burst mode and then continued via a stable release with a low rate. Also, the release profile successfully extended to 21 days, meaning proper time to complete in vitro osteogenic differentiation of stem cells (Bou Assaf et al., 2019) (Figure 5). Of course, the actual burst release is reduced through bonding between HNT and ZA. In addition, the low crystallinity of the scaffolds with more ZA facilitated the water diffusion into the scaffolds and consequently increased the ZA release from the samples. These results are in accordance with the previous study (Karavelidis et al., 2011), which proved that the ZA release rate is amplified by lowering the crystallinity of the polymer matrix.

Cell viability and cytocompatibility of the scaffolds are shown in Figure 6. Human Adipose Stem Cells (hASCs) were cultured onto the samples, and an MTT assay was done. Cell proliferation was investigated after 1, 4, and 7 days. In a study by Davydenko et al., OD in a sample with pure gelatin scaffold was higher than OD of TCPs, which can be considered due to the presence of more sites for cell-matrix interaction. These places are mostly due to the presence of interconnected micron porosity (Davidenko et al., 2016). These results were also observed in this study, which shows the importance of interconnected porous structures. According to the results, both Gel-HNT and Gel-HNT/ZA showed more proliferation than Gel sample during the days of culture, especially in ZA-containing scaffold. By adding HNTs nanorods to the structure, the surface-to-volume ratio increases dramatically, and consequently the cell-matrix interaction increase, which in turn increases cell proliferation and differentiation, which is consistent with previous studies (Ji et al., 2017).

Alkaline phosphatase (ALP) activity assay was used to investigate the osteogenic ability of hASCs. This enzyme is one of the most famous early markers of osteogenesis. There are some researches that demonstrated the role of ALP in osteogenic differentiation (Ardeshirylajimi et al., 2015). The ALP is the final marker for stem cell osteogenic differentiation that promotes the mineralization of calcium (Boraei et al., 2020). This test was performed on scaffolds for 7 and 14 days (Figure 7). The results showed a higher expression of the activity of samples containing ZA on day 14. There was also a slight increase in ALP on day 7. It is important to note that ALP values for all scaffolds were significantly higher than those calculated for TCPs. In addition, the ALP amount is considerably more in the samples including ZA, suggesting the speed up of osteogenic differentiation of seeded hASCs by the scaffold-released ZA. As reported previously, maximum ALP activity is caused by enhancement in the mineralization (Santo et al., 2012).

The state of mineralization, which is the formation of inorganic calcium, was also investigated for 7, 14 and 21 days (Figure 8). This state of mineralization (calcium deposition) acts as the late marker of osteogenesis were investigated by Alizarin Red Staining (ARS) (Veernala et al., 2019). In this analysis, the red dots after imaging indicate the amount of calcium deposition at specific times. The amount of deposited calcium was remarkably increased with increasing the incubation time and adding the ZA until 21st day. On day 14, ZA release from the scaffolds caused a large difference in the amount of calcium deposited. The Bone Mineral Density (BMD) and osteogenic ability of ZA have been widely investigated and different mechanisms have been discussed (Pavón de Paz et al., 2019; Jin et al., 2020). In previous studies on ZA, the rate of increase in calcium deposits was observed with increasing time, the amount of calcium islets on the 14th day was significantly higher than the 7th day, which is consistent with our present study that the deposits Calcium has increased over time and the amount of ZA in the scaffold has increased (Raina et al., 2020; Demir-Oğuz and Ege, 2021).