Osteogenesis capability of three - dimensionally printed poly ( lactic acid )- halloysite nanotube sca ﬀ olds containing strontium ranelate

: In this study, three - dimensional ( 3D ) printing of 3D sca ﬀ olds containing halloysite nanotubes ( HNTs ) and strontium ranelate ( SrR ) as a carrier for the promo tion of bone regeneration is investigated. SrR acts as an anabolic bone - forming and anti - catabolic agent, while HNTs act as a carrier of SrR. Poly ( lactic acid ) ( PLA ) is used as a biodegradable matrix and carrier for HNTs and SrR. The e ﬀ ects of the SrR addition on the morpho -logical, biological, and in vitro release properties of the sca ﬀ olds are evaluated. The morphological results show a homogeneous structure with a proper pore size ( approxi mately 400 µm ) suitable for osteogenesis. The contact angle is decreased after the addition of SrR to the sca ﬀ old to 67.99°, suitable for cell attachment. X - ray di ﬀ raction shows that the SrR is homogenously and molecularly dis tributed in the PLA matrix and reduces the crystallinity in the prepared sca ﬀ olds. The in vitro release results demon strate that the release pro ﬁ le of the SrR is stable, relatively linear, and continuous within 21 days ( 504 h ) . A cumula tive release of SrR of approximately 49% is obtained after a controlled release for 504 h ( 21 days ) and a low primary burst release ( 12% ) . Human adipose stem cells cultured on the 3D - printed sca ﬀ olds demonstrate that the SrR can e ﬃ ciently promote biocompatibility, alkaline phosphatase activity, and alizarin red staining.


Introduction
In general, bone is a composite composition that has three main parts: matrix, fiber, and cell. The most important part is the collagen matrix, which bears the tensile mechanical loads, while the mineral phase, which consists of calcium phosphates, bears the compressive mechanical loads [1][2][3][4]. In this regard, the bone tissue experiences numerous defects throughout life, including problems caused by trauma, injuries of various origins, and aging, which are increasingly studied to address the related problems [5][6][7][8][9]. Considering the importance of these problems, bone tissue engineering is a new progressing method considered to mitigate osteogenesis [10][11][12][13]. Scaffolds are the most important part of tissue engineering science. The scaffold is a matrix in three-dimensional (3D) structures, which has essential characteristics, such as biocompatibility and proper mechanical properties, induces cellular activity and protein production, and provides cell attachment, differentiation, and proliferation [14][15][16][17][18][19]. The interconnecting pore is a significant factor for synthetic scaffolds in bone tissue engineering applications [20,21]. The pore sizes must be around 300 µm for good vascularization, cell attachment, and growth guidance in three dimensions [22]. Several methods are used to obtain porous scaffolds, such as solvent casting, foam gel method, freeze-drying, thermally induced phase separation, particle/salt leaching, and chemical/gas foaming [23][24][25][26][27][28][29]. In recent years, additive manufacturing is widely used in bone repair applications.
Composite structures have been synthesized for various applications, which include ceramic, polymer, and metal compounds [30][31][32][33][34][35][36][37]. Polymer-ceramic nanocomposites with drug-loading properties are widely used in tissue engineering and in cases that require appropriate mechanical properties as well as controlled release of drugs [38,39]. Synthetic biopolymers, including poly(lactic acid) (PLA) and polycaprolactone, are useful for numerous applications in the synthesis of nanocomposites [40][41][42][43][44][45][46][47]. A 3D-printed PLA was investigated in recent years. PLA can be 3D printed at low temperatures, while the agent is bound by a binder solution [48][49][50]. In the 3D-printing method, biocompatible polymers can be used, and thus, composite scaffolds can be designed and manufactured [51]. The osteogenesis agent used in the matrix should be elucidated and its local release should be effective and beneficial. Among such agents, strontium ranelate (SrR), which has anabolic and anticatabolic properties, is increasingly used in tissue engineering scaffolds [52]. The addition of SrR increases the adhesion, proliferation, alkaline phosphatase (ALP) activity, mineralization, and angiogenesis of the 3D-printed scaffolds [53,54]. Another important factor of tissue engineering scaffolds containing drugs is the controlled release of the drug for a proper tissue regeneration. Halloysite nanotubes (HNTs) could help control the release of SrR and reduce its initial burst release [55]. HNT is a natural aluminosilicate that attracts considerable attention and has many applications [56][57][58][59]. HNT has unique properties, such as good mechanical and thermal properties, and is a suitable carrier of drugs. Several studies have been carried out on its applications and properties [60][61][62][63][64].
In this study, the 3D-printing technique was employed to obtain PLA/HNT-SrR scaffolds with different compounds (PLA, PLA/HNT, and PLA/HNT-SrR). The scaffold morphologies were evaluated by scanning electron microscopy (SEM). An X-ray diffraction (XRD) analysis was performed to elucidate the crystallographic features of the scaffolds. The drug release behavior of the SrR-loaded scaffolds was investigated. In addition, the cellular behaviors 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), ALP activity, and calcium assay of the 3D scaffolds were revealed. Suspensions of HNT-and SrR-modified HNT dispersed by adding 2 wt% CTAB (with respect to the solid content of HNT) in a PLA solution in THF (80 g/L) were prepared and granulated as described elsewhere [55,65]. Granules were crushed into fine powders for the extrusion of colloidal filaments and printing [66]. The 3D scaffold was obtained by a 3D-printing device (Prusa I3 with the Repetier software). 3D scaffolds with different compositions (PLA, PLA/HNT, and PLA/HNT-SrR) were printed (diameter: 1 cm, height: 0.2 cm, infill density: 40%, four layers). The prepared scaffolds were immersed in NaOH (20 wt%) for 5-10 s for a cross-linking treatment and then washed with deionized H 2 O.
The PLA-based scaffolds were coated with gold (SC7620, QUOROMTECHNOLOGIES-EMITECH, England), and then, the microstructure was observed by SEM (AIS2100, SERON TECHNOLOGY, South Korea) in the secondary-electron mode at 20 kV to evaluate the morphology. Energy-dispersive spectrometry (EDS) was used to confirm the presence of HNTs and SrR in the scaffolds as well as their uniform distributions.
An attenuated total reflectance (ATR) Fourier-transform infrared (FTIR) analysis (Spectrum 100, PerkinElmer Company, UK) was used to investigate the functional groups and structural composition of the scaffolds.
The crystallography structures of the scaffolds were evaluated by XRD (EQUINOX3000, Inel, France) with Cu Kα radiation at a voltage of 40 kV and a scan rate of 2°/min.
The hydrophilicity of the scaffolds was evaluated by a contact angle analysis. Approximately 20 µL of distilled water was dropped on a flat surface part of the scaffold. The contact angle was recorded after 2 s. A drop shape analysis software was applied to determine the baseline and contact angle. The results are presented as mean and standard deviation of five replications for each sample.
The release profile of SrR from the SrR-containing scaffold (0.5 g) was investigated by immersing the scaffold in 5 mL of PBS. It was then incubated in a shaker incubator at 90 rpm. At selected time points, 2 mL of the PBS was picked up and replaced with 2 mL of fresh PBS. The amount of released SrR was evaluated by an ultraviolet (UV)-visible device (NANODROP 2000c, Thermo Scientific Co., USA) at several time points (24,48,72,120,168,336, and 504 h) at a wavelength of 318 nm.

Cellular assay
The prepared scaffolds were punched, placed into tissue culture polystyrene (TCPS), and then sterilized by 70% ethanol for 2 h and 20 min UV irradiation on each side. After immersing the scaffolds in a culture medium overnight, the density of primitive cells was 2 × 10 5 cells/cm 2 . They were suspended in 200 μL of HG-DMEM and then seeded on the prepared scaffolds and controls (TCPS). The incubation was carried out for 30 min. Afterward, the basal medium (800 μL) was poured into the wells. All scaffolds were moved to new plates. The osteogenic medium was added and kept for 7 and 14 days. The renovation of the osteogenic medium was carried out every 2 days.
The cytotoxicity and viability of human adiposederived stem cells (hASCs) on the prepared scaffolds were investigated by an MTT assay. The scaffolds were punched and placed into the plates. The primitive density of hASCs was 2 × 10 5 cells/cm 2 . The plates were then placed into the incubator. Fifty microliters of the MTT solution (5 mg/mL) were then poured into the wells. The medium was removed after 2 h from the incubation. Three hundred microliters of dimethylsulfoxide (Merck) were then replaced to dissolute the dark-blue intracellular formazan. This procedure was carried out at the first, fourth, and seventh days of cell seeding. The dye solutions were then moved to six-well plates to observe the optical density of each well at 570 nm by a spectrophotometer (BioTek Instruments, USA).
Radioimmunoprecipitation assay buffer (200 μL) was used to extract the total protein from hASCs cultured on samples and TCPS for ALP activity evaluation at the time points of 7 and 14 days. The lysate was centrifuged (1,200 rpm, 4°C, 5 min) for the sedimentation of cell debris. An ALP assay kit (Parsazmun Co., Tehran, Iran) was used to measure the ALP activity of the collected supernatant.
The cresolphthalein complexone method was used to evaluate the amount of deposited calcium on the samples and TCPS. The homogenization of the hASCs was carried out by HCl (0.6 N, Merck Co.). They were shaken for 4 h at 4°C. In this stage, after the addition of a reagent to the calcium solution, the optical density was measured at 405 nm [38].
The mentioned analyses were carried out three times. Mean ± standard deviation was used to present the final data. The differences between the results were evaluated by a one-way analysis of variance. All results were statistically investigated at a level of p < 0.05.

Characterization of the 3D-printed scaffolds
A proper scaffold provides specified features, such as a porous structure and interconnected pores for cell adhesion, proliferation, and differentiation, as well as transfer of nutrients to the cells and removal of waste products from cells [38]. Figure 1 shows SEM images of the 3Dprinted scaffolds and macroporous structure. The homogeneous porous and completely interconnected matrix were detected in the scaffolds. According to the fixed setup of the printer, all samples have the same pore size (the mean value is 400 µm), which is in good agreement with other studies and in the suitable range for bone regeneration applications [67,68]. The increase in HNT content reduces the smoothness of the scaffold surface, mainly owing to the tendency of the nanoparticles for agglomeration at high concentrations [69]. HNT is stiffer than molten PLA, and thus, the polymer drags partially with the nozzle movement during the 3D printing [70].
To confirm the presence of HNTs and SrR in the scaffolds, an EDS analysis of Al, Si, and Sr elements was carried out on the scaffolds, as shown in Figure 2. The images show the uniform distribution of HNTs (Al and Si elements) and SrR (Sr element) in the PLA matrix. Figure 3 shows the results of the ATR FTIR spectroscopy of the scaffolds. The typical bands of PLA (polyesters) are observed, including those at 2,921 and 2,851 cm −1 attributed to symmetric and asymmetric vibrations of the C-H bond of the CH 3 groups, respectively, at 1,741 and 1,179 cm −1 related to C]O and C-O-C stretching, respectively, at 1,450 cm −1 corresponding to C-H stretching vibration in methyl groups, at 1,363 cm −1 assigned to symmetric vibrations of -CHbending, at 1,076 cm −1 related to carbonyl C]O and -OH groups, and at 867 and 753 cm −1 assigned to the C-C stretching vibration in the structure [71,72]. After the addition of HNTs to the structure, the bands at 2,851 and 2,925 cm −1 attributed to PLA slightly shifted to higher wavenumbers, which indicates hydrogen bonding of the hydroxyl chains of PLA with the siloxane groups of halloysites. The band at 1,023 cm −1 corresponds to the stretching of the Si-O group, which indicates the successful incorporation of the HNT in the structure [73]. After the loading of SrR, bands at 1,260 and 1,355 cm −1 appeared, which are related to the C-N vibration in SrR. This indicates a successful loading of the SrR drug into the scaffold structure. Figure 4 shows XRD results of the prepared scaffolds. In general, they have two characteristic peaks at 2θ of 17°a nd 19° [74]. A similar peak is observed in our results (2θ = 16.72°) but is too broad, which indicates that the PLA is amorphous during the scaffold preparation. The HNTs usually have several specific peaks. In this study, owing to the low percentage of HNTs in the structure of the scaffolds, not all of their peaks are clear. Only the characteristic peak at 2θ = 12.05°for the PLA/HNT scaffold, which contains HNTs, is observed. This peak is similar to the major peak of HNTs at 2θ = 11.9° [75]. The peak demonstrates the presence of HNTs in the structure. This peak is in good agreement with the (001) basal plane of HNTs [68]. After the addition of SrR to the structure, no characteristic peaks of SrR were observed owing to two reasons: (1) low percentage of SrR in the scaffold, which makes it undetectable for the device, and (2) with the addition of SrR to the polyesters and polyethers that compose the scaffold matrix, the crystallinity of the SrR is reduced, and thus, it tends to become amorphous [39].
The contact angle test results of the PLA, PLA/HNT, and PLA/HNT-SrR scaffolds are presented in Figure 5. As expected, the contact angle of the PLA scaffold was the largest, 111.1 ± 13.63°, which is in the range reported for polyesters [76]. Upon the addition of HNTs to the    structure, the contact angle and hydrophobicity of the structure are reduced (100.15 ± 11.23°). After the addition of SrR to the structure, which is a completely hydrophilic drug, the hydrophobicity of the structure was reduced and the angle reached 67.99 ± 5.47°. In general, with the increase in the amount of hydrophilic components in the structure (in this case, SrR), a decrease in contact angle is expected. The suitable contact angle range for the hydrophilicity of scaffolds and cell attachment to the scaffold is 40-70°, considered an important parameter [39,77]. Figure 6 shows the release profiles of SrR from the scaffolds. The HNT was added as a drug carrier to the scaffold to control the release behavior of the SrR. Twenty-one days (500 h) is a suitable time for osteogenesis differentiation by continuous release of SrR. The porous 3D scaffold exhibited a relatively linear and stable release behavior of SrR. The cumulative release after 500 h was approximately 49%, similar to our previous result [38].

In vitro SrR release investigations
Owing to the formation of SrR bonds on or into the HNT, the SrR release profile exhibits a controlled behavior. As shown in Figure 6, the initial burst release rate of SrR from the scaffold is extremely low, approximately 12%. The initial burst release is very high in many systems, particularly in systems without drug carriers. In general, two mechanisms can be proposed for the release of SrR from HNT-containing nanocomposite scaffolds. The first mechanism is attributed to siloxane groups (Si-O-Si) on the HNT surface and their negative charge. Thus, they can bond the positively charged strontium. The second mechanism is attributed to aluminum hydroxide groups (Al(OH) 2 ) inside the positively charged HNT lumens. Thus, there is a possibility of bond formation between them and the negative terminal of the ranelate [55].

Cellular analysis
Biocompatibility and nontoxicity are important factors for a good scaffold. An MTT assay was performed to investigate the cytocompatibility of the prepared samples by culturing hASCs on them (Figure 7). The proliferation of hASCs on the SrR-containing scaffold after 1, 4, and 7 days was better than those for the PLA and PLA/HNT samples. After 5 days, it was better in the case with the SrR scaffold than for the other samples. The extracts acquired from the PLA/HNT and PLA/HNT-SrR scaffolds exhibited superior cell viabilities after 1, 4, and 7 days [78]. Thus, all prepared scaffolds in our experiment exhibited a good cytocompatibility and were suitable for bone tissue engineering. The results of the MTT analysis can be explained using the EDS analysis, where the scaffold containing the drug and element strontium exhibited a higher cell proliferation, as expected.
The ALP activity was evaluated to investigate the osteogenic potential of the cells. ALP is an important osteogenic marker expressed in the early stages. The results showed an increase in the ALP activity. Despite the release of SrR, it does not have a significant effect on the hASCs growth in the first week. There is a notable difference between the SrR-containing scaffolds and the  rest of scaffolds in the second week [38]. Figure 8 shows a higher ALP activity in the PLA/HNT-SrR scaffold than in the PLA and PLA/HNT scaffolds at day 14 (p < 0.05). These results are consistent with the contact angle analysis, where the scaffold containing SrR had the optimal contact angle and consequently higher hydrophilicity, which, in turn, increased the cell attachment and improved the APL results.
In general, large amounts of calcium deposits are observed in bone tissue. Alizarin red staining (ARS) is one of the approved evaluation methods for calcium deposition and subsequent mineralization of synthesized scaffolds. In this study, such a test was performed after 7 and 14 days. Figure 9 shows an increase in the calcium content after 14 days of culturing, higher than that after 7 days. The scaffold containing SrR exhibited the highest amount on day 7 and the highest increase on day 14. The SrR-containing scaffold had the largest amount of deposits among the prepared scaffolds [79]. As shown in Figure 9, the ARS quantification indicates that the deposition of calcium in the SrR porous scaffold was considerably higher than that in the neat scaffold. Hence, the SrR-containing scaffold accelerates the bone formation-enhanced osteogenic differentiation. Several mechanisms about the osteogenesis ability of SrR have been investigated [80][81][82].

Conclusion
The 3D-printing technology was used to produce proper scaffolds for bone tissue engineering applications. Interconnected porous structures with pore sizes of approximately 400 μm and high porosities were produced. The addition of SrR to the scaffolds remarkably changed the properties of the 3D-printed scaffolds. A cumulative release of SrR of approximately 49% was obtained after the controlled release for 504 h (21 days) and low primary burst release (12%). The sustained release of SrR could be observed for at least 504 h (21 days) owing to the loading of SrR on or into the HNTs. The contact angle largely decreased upon the   addition of SrR to the structure to 67.99°in a suitable range for cell attachment. The nontoxicity of HNT and SrR was demonstrated after their addition to the PLA scaffold. The cellular analysis demonstrated that the ALP activity and calcium deposition increased upon the addition of SrR to the scaffold. Thus, the produced scaffolds can provide a good porous 3D structure in bone tissue engineering applications and prompt bone regeneration. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.

Conflict of interest:
The authors state no conflict of interest.