Reduced graphene oxide: osteogenic potential for bone tissue engineering

Abstract

Collagen (Col) type I, as the major component of the bone extracellular matrix has been broadly studied for bone tissue engineering. However,inferior mechanical properties limit its usage for load bearing applications. In this research, freeze dried Col scaffolds are coated with graphene oxide (GO) through a covalent bond of the amine Col with the graphene carboxyl groups. The prepared scaffolds were then reduced using a chemical agent. Scanning electron microscopy exhibited a porous structure for the synthesized scaffolds with an approximate pore size of 100–220 ± 12 µm, which is in the suitable range for bone tissue engineering application. Reducing the GO coating improved the compressive modulus of the Col from 250 to 970 kPa. Apatite formation was also indicated by immersing the scaffolds in simulated body fluid after five days. The cytocompatibility of the scaffolds, using human bone marrow‐derived mesenchymal stem cells, was confirmed with MTT analysis. Alkaline phosphatase assay revealed that reducing the Col–GO scaffolds can effectively activate the differentiation of hBM‐MSCs into osteoblasts after 14 days, even without the addition of an osteogenic differentiation medium. The results of this study highlight that GO and its reduced form have considerable potential as bone substitutes for orthopaedic and dental applications.

1 Introduction

One of the major challenges in bone tissue engineering is the design of a matrix that can imitate bone tissue properties to provide a temporary bone regeneration scaffold that combines growth factor, stable mechanical properties, and absorbable implant as an alternative for a natural bone structure [1, 2]. If these parameters are optimised to provide the required biomechanical properties during tissue regeneration, the synthesised structure can function in accordance with the natural properties of the bone tissue. Furthermore, recent studies on tissue engineering have focused on designing and implementing scaffolds that can provide a passive structure to support bone cells as well as stimulate the differentiation and proliferation of osteoblasts [3, 4]. In recent years, scientists in tissue engineering have focused on designing and building materials and eventually scaffolds with a bone‐like structure. To achieve this goal, a nanocomposite system containing both organic and inorganic bone components is required [5]. Techniques based on biomimetic strategies have been applied to design bone scaffolds that can mimic the natural bone structure and its chemical composition. A wide range of natural polymers including gelatine [6], collagen (Col) [7], chitosan [8], and hydroxyapatite [9] have been applied as biomimetic materials owing to their significant advantages such as biocompatibility and bioactivity. Among these, Col, the main component of the bone extracellular matrix (ECM), is known as an excellent biocompatible polymer that exhibits acceptable biodegradability providing binding sites suitable for cell attachment [10]. However, this biopolymer lacks adequate stiffness to be used for load bearing applications [11].

A wide range of research has been focused on carbon‐based nanostructures including zero‐dimensional (0D), 1D, 2D, and 3D types, as reinforcing agents [12, 13, 14]. Recently, owing to its promising physicochemical properties, graphene oxide (GO) has demonstrated excellent potential for biomedical applications such as drug/gene delivery [15], bioimaging [16], and tissue engineering [17]. Furthermore, graphene and its derivatives have been considered as an attractive component for stem cell cultures owing to their significant effects on enhancing stem cell proliferation, differentiation, as well as promoting cell adhesion and growth. One of the considerable features of GO compared to other carbon‐based nanostructures is its high mechanical stiffness which, in addition to bioactivity, makes it a remarkable candidate for hard tissue regeneration [18, 19, 20]. Improved biomimetic mineralisation and enhanced therapeutic effects for repairing the cranial defect after GO functionalisation on Col scaffolds have been previously confirmed [21]. Radunovic et al. demonstrated that GO coating on Col membranes promotes osteogenic differentiation of human dental pulp stem cells (DPSCs). Reduced inflammatory markers of tumour necrosis factor α and cyclooxygenase‐2 on DPSCs for GO‐containing membranes were observed with respect to those uncoated [22].

In this study, porous Col scaffolds were fabricated via a freeze‐drying method and coated chemically with GO followed by chemical reduction to obtain Col‐reduced GO (rGO) samples. Human bone marrow‐derived mesenchymal stem cells (hBM‐MSCs) were then cultured in scaffolds and the osteogenic capability of the rGO was assessed by investigating alkaline phosphatase (ALP) activity after 7 and 14 days.

2 Materials and methods

2.1 Materials

Graphite powder, sodium nitrate, potassium permanganate, N ‐ethylcarbodiimide hydrochloride (EDC), N ‐hydroxysuccinimide (NHS), phosphate buffered saline (PBS), gentamicin, penicillin–streptomycin, 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT), dimethyl sulphoxide (DMSO) and sodium hydrosulphite were purchased from Sigma‐Aldrich Chemie GmbH (Germany). Sulphuric acid (H₂ SO₄), hydrogen peroxide, acetic acid, sodium hydroxide, hydrochloric acid (HCl) and glutaraldehyde were obtained from Merck (Germany). Ethanol 99% was purchased from Pars Alcohol Company (Iran). Col type I was obtained from Nano‐Zist Arrayeh Company (Iran). Foetal bovine serum (FBS) was purchased from Gibco (Germany). Dulbecco's modified Eagle medium (DMEM)‐F12 and trypsin‐ethylenediaminetetraacetic acid were purchased from Invitrogen (Germany). All other reagents or solvents were of high‐performance liquid chromatography grade.

2.2 Synthesis of GO

GO flakes were synthesised via a slightly modified Hummers method [23]. Briefly, 0.5 g of graphite powder, 0.5 g of sodium nitrate, and concentrated H₂ SO₄, 25 ml were placed into a 250 ml round bottom flask cooled at 0 °C and this mixture was stirred for 30 min for homogenisation. Then, 3.5 g of potassium permanganate was added to the reaction system gradually over 1 h. The temperature of the mixture was retained about 10°C. The temperature of the reaction system was then increased to 35°C and stirred for 2 h. Afterwards, 45 ml of deionised (DI) water was slowly added into the paste‐like product. Finally, 1.5 ml of 30% hydrogen peroxide and 27.5 ml of DI water were poured into the reaction system resulting in the formation of yellowish brown suspension. The solid product, graphite oxide, was separated from the reaction mixture by centrifugation (at 5000 rpm for 30 min). The yellowish brown solid powders were washed for three times with diluted HCl (3%), subsequently, this product was rinsed by DI water several times to eliminate salt residues and acid completely, and was kept dry in a desiccator for further experiments.

2.3 Synthesis of Col scaffolds

To prepare a Col porous 3D scaffold, first, type I Col was dissolved at a concentration of 10 mg/ml in 1% acetic acid, being placed on the magnetic stirrer for 20 h to get dissolved well. The homogenous suspension was poured in Teflon moulds and frozen at a temperature of −20°C for 8 h and −80°C for 12 h. Then, the frozen mixture was put in a freeze dryer device (Alpha 1–4 LDplus, Martin Christ) in −50°C for 24 h. To obtain the cross‐linked scaffolds, samples were immersed in ethanol (90% v/v) solution consisting of EDC (0.0024 g) and NHS (0.0012 g) overnight. Then, samples were washed with DI water several times and finally lyophilised for another 24 h.

2.4 Col scaffold coating and reduction

The coating was performed according to our previous procedure [24]. Briefly, a solution of GO was prepared in distilled water at a concentration of 100 μg/ml being well stirred on an ice bath with a sonication device for 30 min. Then, the solution containing water, GO and EDC was prepared after 15 min stirring with a ratio of 5:4 for GO:EDC, respectively. Col scaffolds were then immersed in the prepared solution and shacked at room temperature for 6 h. Then, the solution was adjusted to pH 4.7 using 0.1 M HCl. After removing the solution, the remaining salt residues were washed with water, and the scaffolds were finally placed in freezing dryer for 24 h. Scaffolds were then reduced by immersing in a solution of sodium hydroxide 2% for 3 min followed by washing with ethanol 90% several times.

3 Characterisations

3.1 GO analyses

Thickness and dimension of the synthesised GO were measured by atomic force microscopy (AFM; Dualscope/Rasterscope C26, DME). Compositional characteristics of the flakes were also assessed by using an X‐ray diffractometer (XRD; Rigaku D/max‐3C), Fourier transform infrared (FT‐IR) spectrophotometer (Bruker IFS 48 instrument) and Fourier transform (FT)‐Raman spectrometer (Senterra, Bruker).

3.2 Scaffold analyses

XRD was performed to detect the phases formed in the scaffolds using a diffractometer with Cu‐Kα (λ = 0.154 nm) radiation in an interval of 2θ range of 5–50° at a scan speed of 2°/min with a step size of 0.02°. This instrument worked with the voltage and current settings of 40 kV and 40 mA, respectively. FT‐IR analysis was carried out by using a spectrometer in the range 500–4000 cm⁻¹ at a scan speed of 23 scans/min with a resolution of 4 cm⁻¹ to investigate the interactions. Raman spectrum was recorded in a scanning range of 800–2200 cm⁻¹. After gold coating, the morphology of the prepared samples was examined by scanning electron microscopy (SEM; Tescan Vega II). The porosity of the scaffold was measured using the liquid displacement method and ethanol was as a liquid medium based on (1)

$$\mathrm{Porosity}\left(\text{\%}\right)=\left(W_{\mathrm{t}}-W_{\mathrm{r}}-W_{\mathrm{d}}\right)/\left(W_{\mathrm{c}}-W_{\mathrm{r}}\right)\times 100,$$ (1)

where W _c is the weight of ethanol containing bottle; W _t is the weight of the ethanol containing bottle after immersion of the scaffold; W _r is the weight of the ethanol containing bottle after removal of the scaffold; and W _d is the weight of the dry scaffold. The evaluation of the scaffold hydrophilicity was also performed with the static water contact angle measurement system (OCA10, Dataphysics) following pouring a 4‐μl droplet of DI water to each surface. All measurements were performed three times. To investigate mechanical stiffness of the scaffolds, the compressive test was performed using a conventional testing machine (H10KS; Hounsfield) at a loading rate of 1 mm/min. The compressive modulus was calculated from the slope of the stress–strain curve in the linear region. The bioactivity of the Col–rGO scaffolds was evaluated by immersing samples in simulated body fluid (SBF) at 37°C for 5 days. After removing the scaffolds from the SBF, they were washed with DI water and dried. The morphology of the apatite crystals and the Ca/P ratio formed on the scaffold surface were investigated using a scanning electron microscope equipped with an energy‐dispersive X‐ray analyser (EDX; Rontec).

3.3 Cytotoxicity and morphology assessments

hBM‐MSCs were purchased from Royan Stem Cell Bank, Royan Institute, Iran. Cells were cultured in DMEM‐F12 supplemented with 100 IU/ml penicillin, 100 IU/ml streptomycin and 10% FBS. Before cell culture, scaffolds were sterilised by utilising ethanol (70% v/v) followed by washing with PBS several times. After that, 7 × 10³ cells were suspended in expansion medium and seeded over scaffolds in 96‐well culture plates being incubated at 37°C in 5% CO₂. The cytocompatibility assessment was performed by MTT colorimetric assay after 24 and 48 h. Briefly, 100 ml MTT solution (5 mg/ml in PBS) was poured into each well, and the scaffolds/cells were incubated for 4 h. Then, the medium was removed and the formazan precipitates were dissolved in DMSO. The optical absorbance at 490 nm was measured using a microplate reader (ELISA reader, ELX808, BioTek). Cells cultured on tissue culture plate (TCP) were considered as the control group. Cell morphology on scaffolds was also assessed by SEM. Briefly, after 48 h, scaffolds were washed with PBS and cells were fixed in 2.5% glutaraldehyde for1 h followed by washing again with PBS several times. Then, the scaffolds were dehydrated by alcohol gradients (30, 50, 70, 90, and 100%) and coated with gold for SEM investigation.

3.4 ALP activity

To assess the osteogenic‐inducing capability of GO coating, ALP activity was measured for hMSCs cultured on scaffolds after 7 and 14 days. First, cell seeded constructs were washed and homogenised with PBS (pH 7.4) followed by sonication. After that, the mixture of cell lysate (0.1 ml) with 0.2 ml of p ‐nitrophenyl phosphate (pNPP) substrate solution (BioVision) was prepared. In the next step, the solution was incubated at 37°C for 30 min and then 2 M NaOH was added to stop the reaction [11]. Absorption at 405 nm was detected using a microplate reader (Stat Fax 3200; Awareness Technology).

4 Statistical analysis

Values were demonstrated as means ± standard deviations (SDs). One‐way analysis of variance using SPSS 16.0 software was used to analyse the data. p  < 0.05 was considered significant. All results were recorded as mean ±SD.

5 Results and discussion

Carbon‐derived nanomaterials have been widely used in bone tissue engineering [25, 26, 27]. Among this family, GO and rGO with exceptional characteristics and biocompatibility have also been broadly used for regenerative medicine [28]. The covalent conjugation of 5 μg/ml of a GO flake solution to Col scaffolds has been previously reported for osteogenic differentiation of hMSCs [29]. In this research, we increased the concentration of GO coating to 100 μg/ml on the Col scaffolds with the aim of assessing the osteogenesis capacity of GO and its reduced form on hBM‐MSCs.

As represented in the AFM images (Fig. 1), the lateral dimension of the synthesised GO flakes was <1 µm and the thickness of the nanoparticles was ∼2 nm, indicating a single layer of GO [22]. It was reported that in the same dosage, nano‐sized GO flakes are more biocompatible and induce less inflammatory responses for cells compared to micro‐sized GO flakes [30]. The porosity, pore size, and morphology obtained by the freeze‐drying method depend on different processing factors including freezing temperature, solution concentration, and solvent properties [31]. As can be observed in the SEM images (Fig. 2), both scaffolds contained porous structures with randomised interconnected pores. Clearly, no difference was observed in the structure of the Col following GO coating. The relative pore diameters of the scaffolds were calculated to be in the range of 100–220 ± 12 µm, which seems to be in a suitable range for the growth and regeneration of bone tissue [32]. The approximate porosity determined by the Archimedean principle indicated a porosity of ∼98.4 ± 0.7% for both scaffolds. Importantly, a porosity of up to 90% is essential for the transfer of nutrients and excretions of the cell, as well as for allowing the migration and penetration of the cells into the bone marrow [33]. Therefore, the porosity of the synthesised scaffolds was considered to be in an acceptable range for the application of bone tissue engineering [33, 34]. Interestingly, the GO coating and reduction process did not affect the structural morphology of the Col scaffolds.

Fig. 1.

AFM image of synthesised GO flakes

Fig. 2.

SEM images of synthesised scaffolds

(a) Col,

(b) Col–rGO. Scale bar = 100 μm

Fig. 3 displays the results of the FT‐IR spectra of Col and Col–GO scaffolds with its reduced counterpart sample. The distinctive features of GO in the FT‐IR spectra can be attributed to the C=O carbonyl stretching at 1720 cm⁻¹, the C–OH stretching at 1210 cm⁻¹, and the C–O stretching at 1040 cm⁻¹ [35]. The spectra also indicated a C=C peak at 1610 cm⁻¹ corresponding to the remaining sp² character [35] and a broad intense peak at 3140 cm⁻¹ for the OH stretching frequencies. The peak at 1350 cm⁻¹ is characteristic for the C–OH stretching vibration frequency of GO [36]. For the Col scaffolds, the absorption of type 1, type 2, type 3, and type A amide groups is visible at 1658, 1540, 1240, and 3396 cm⁻¹, respectively. In the results obtained for the Col–rGO‐100 scaffolds, the absorption of a carboxylic group at 1040 cm⁻¹ and carbonyl group at 1728 cm⁻¹ was not observed, which can be attributed to the reduction of the carboxylic component of graphene sheets after bonding with the amine groups of the Col and the chemical reduction of the scaffolds [37]. A weak peak was also observed at 1447 cm⁻¹, which is related to C–OH stretching in the Col–rGO‐100 scaffold, which was sharper in the GO spectra. Col amide peaks were also adsorbed at 1237, 1546, 1661, and 3334 cm⁻¹ for Col–rGO‐100. Therefore, according to the results of the FT‐IR spectrum, the chemical bonding between the GO and Col was confirmed. It should also be mentioned that the reduction process was accomplished successfully. Fig. 4 also presents Raman spectrum for the Col–rGO sample representing the D band at 1355 cm⁻¹, which is associated with the sp³ vibrations of carbon atoms and also a G band at 1529 cm⁻¹ attributed to the sp² hybridised carbon atom of GO [27].

Fig. 3.

FT‐IR spectra of

(a) GO,

(b) Col,

(c) Col–rGO

Fig. 4.

Raman spectroscopy of Col–rGO scaffold

The XRD pattern of the synthesised GO, Col, and Col–rGO scaffold are displayed in Fig. 5. A sharp peak at 2θ  = 11.3° corresponds to the d ‐spacing of 0.84 nm in GO [38]. At 2θ  = 8°, the diffraction represents a Col equatorial structure. Furthermore, the diffraction at 2θ  = 21° corresponds to the distance between the Col triple chain spiral structures [39]. For the Col–GO scaffold, apart from attributing Col peaks at 2θ  = 8° and 21°, the lack of regular GO diffractions could correspond to the fully exfoliated structure formation and homogenous dispersion of the GO within the Col matrix [27]. Contact angle analysis was used to calculate the hydrophilicity of the synthesised scaffolds. Hydrophilicity is one of the most important factors influencing the interaction between the surface of the scaffolds and the cells. The contact angle of the Col scaffold was recorded to be 111.7° ± 3.03°. As expected, the Col–GO and Col–rGO scaffolds contact angles were decreased to 70.5°±2.75° and 81.9°±2.93°, respectively. This increase in the hydrophilicity of the scaffolds is due to the presence of hydroxyl and negative charge carboxyl groups in the structure of the GO. Moreover, it is reported that rGO exhibits less hydrophilicity compared to GO [15]. According to previous studies, scaffolds with a water contact angle in the range of 40–80° have been demonstrated to exhibit acceptable affinity for supporting adhesion of different cell types, it can be concluded that the synthesised scaffolds have suitable surface hydrophilicity for interaction with the cell [40]. To investigate the effect of the GO coating on the mechanical properties of the scaffolds, the compressive modulus was calculated from the linear region of the strain stress diagram (Fig. 6). With the addition of the GO, because of the strength and high modulus of this nanomaterial, the compressive modulus of the Col scaffold was increased significantly from 250 ± 7 to 420 ± 11 kPa (p  < 0.001). Furthermore, the same trend was observed after reducing the scaffolds, the compressive modulus was increased significantly to virtually 600 ± 16 kPa (p  < 0.001). This increase is presumably due to the distribution of graphene particles within the scaffold as well as the increase in the thickness of the GO layers following the reduction process [25]. By increasing the thickness of the GO layers, the gap between the layers decreases and the van der Waals gravitational force is created between the single‐layer of the GO coating [25].

Fig. 5.

XRD patterns of

(a) Col–rGO,

(b) GO,

(c) Col

Fig. 6.

Mechanical analysis of the scaffolds

(a)–(c) Stress–strain curves,

(d) Compressive modulus of the scaffolds. **Significant increase for Col–rGO compared to other samples, and for Col–GO compared to Col scaffold (p  < 0.001)

Apatite formation on the surface of the Col–rGO scaffolds immersed in SBF was used to predict the potential of bone‐forming in vivo. Morphological changes indicating a crystalline layer of hydroxyapatite formation on the surface of the scaffold was observed in the SEM images (Fig. 7). EDX analysis indicated the atomic ratio of Ca/P to be about 1.61 after five days, which agreed with the hydroxyapatite stoichiometric ratio.

Fig. 7.

SEM image of

(a,b) Col–rGO scaffold after immersion in SBF,

(c) Energy‐dispersive X‐ray spectroscopy spectra of scaffold exhibiting the peaks of Ca and P

The results of the MTT test indicated that the cell viability of the scaffolds was virtually the same among all samples after 24 and 48 h (p  > 0.05) (Fig. 8). No significant differences between the viability of the hMSCs cultured on the scaffolds and those on the TCP were observed. Although not significant, the viability of the hMSCs reduced when cultured on Col–rGO scaffolds compared with that on TCP. This finding could be due to the presence of reactive oxygen species (ROS), which can subsequently increase oxidative stress and reduce biocompatibility. ROS inhibits the cellular antioxidant enzyme functions, which leads to the generation of free radicals, which can damage the DNA as well as the lipid membrane of the cells [41]. The production of ROS by rGO is less than that of GO. Thus, the number of viable cells in the rGO sample could be greater than those of GO [42]. Conversely, it has been noted that adding GO into Col enhances the roughness of the scaffold providing an appropriate environment for mesenchymal stem cells (MSCs) to adhere [43]. High surface area, π–π interactions, and hydrophilicity of GO functional groups could improve the surface protein adsorption from the surrounding, leading to a desirable microenvironment for MSC adhesion and proliferation [44]. SEM images of hMSCs seeded on the Col and Col–rGO scaffolds are displayed in Fig. 9. As can be observed, both scaffolds supported cell adhesion with Col–rGO providing a possibly more desirable environment for cells to spread and flatten.

Fig. 8.

Viability of hMSCs cultured on prepared scaffolds. There was no significant toxic effect among samples compared to the control group after 24 and 48 h (p > 0.05). Data were presented as mean ± SD from three independent experiments

Fig. 9.

SEM micrograph of hMSCs cultured on

(a) Col,

(b) Col–rGO scaffolds after 48 h. Scale bar = 100 μm

The expression of ALP was determined by pNPP assay after 7 and 14 days to assess the functional activity of the hBM‐MSCs on the prepared scaffolds. As can be observed in Fig. 10, at day 7, the ALP activity in the Col–rGO and Col–GO samples was significantly greater than that for the other samples. The same trend was also observed after 14 days with higher activity of ALP for GO containing scaffolds. Although after 14 days the Col–rGO had the highest ALP activity, this number was not significant compared to Col–GO scaffold. The obtained results could be due to the different physicochemical properties of GO and rGO. It is reported that the incorporation of GO into natural polymers such as Col and chitosan can enhance the number of functional groups within the polymer structure and help to mimic more precisely the bone tissue features and ion organisation to mineralise scaffolds [45].‏ GO and rGO demonstrated a high level of calcium absorption in the medium. This amount of absorption is due to the oxidation factors on the surface of these materials. The graphene hydroxyl groups on the scaffold surface cause hydrogen bonding between the proteins and the surface of the scaffold. Furthermore, the greater absorption of calcium by the rGO has been attributed to an increase in the thickness and distance between the lower layers [46].

Fig. 10.

Comparison of the ALP activity for the hMSCs cultured on prepared scaffolds after 7 and 14 days. *Significant higher ALP activity was observed for both Col–GO and Col–rGO compared with other samples (p < 0.05). Data were presented as the mean ± SD (n = 3)

A recent study has indicated that GO can induce prostaglandin E2 (PGE2) secretion, which is known as an important mediator of osteoblastic differentiation. PGE2 is crucial for the up‐regulation of transforming growth factor‐β1, where both facilitate differentiation by increasing the ALP specific activity [47]. The gene expression analysis of osteogenic proteins has demonstrated that bone morphogenetic protein 2 as a bone formation stimulator [48] has been up‐regulated on high concentration GO‐coated substrates. Furthermore, GO has been demonstrated to increase the expression of runt related transcription factor 2, an important mediator for osteoblast differentiation, maturation, and homeostasis [49]. Another reason that can be attributed to enhancing osteogenic differentiation and the activity of ALP is the improved mechanical properties of the scaffolds by adding GO [50]. Biophysical factors such as mechanical interactions between cells and their extracellular environment can strongly influence stem cell behaviour [51]. It has been demonstrated that stem cell behaviours can be influenced by mechanical forces as biophysical signalling that can control stem cell self‐renewal and differentiation [52, 53, 54]. It is noted that stiffness of the matrix can induce the cellular mechano‐transduction cascade initiation, which is an accelerator of cell differentiation. These effects are implemented via intracellular signalling pathways that transform force cues into biochemical signals that trigger fundamental cellular processes such as cell adhesion, proliferation, and differentiation [55, 56]. It has been demonstrated that MSCs cultured on extracellular matrices with different stiffness under similar medium conditions exhibit different results. Very soft substrates (0.1–1 kPa) induce neurogenic differentiation, stiffer (8–17 kPa) substrates cause muscle formation, whereas the stiffest (25–40 kPa) substrates promote bone cells [57], i.e. MSCs seem to differentiate into tissues with stiffness close to that of the basal ECM. It is also reported that the natural bone compressive modulus (Osteoid) range, which is reported to be >30 kPa, could be used to differentiate MSCs to osteoblasts and could be applicable in bone tissue engineering [29]. Moreover, recent studies have proven that a rigid matrix that can mimic the pre‐mineralised bone matrix promotes the MSCs osteogenic differentiation [29].

6 Conclusion

We demonstrated that GO coating can effectively improve the osteogenic capability of Col scaffold. Although reducing significantly improved the mechanical properties of the GO, this result was not significant in terms of ALP expression. Considered together, both GO and rGO could provide a promising platform for bone tissue engineering with the rGO‐containing scaffold being more appropriate for a load bearing application.