Treatment of steroid-induced osteonecrosis of the femoral head using porous Se@SiO2 nanocomposites to suppress reactive oxygen species

Reducing oxidative stress (ROS) have been demonstrated effective for steroid-induced osteonecrosis of the femoral head (steroid-induced ONFH). Selenium (Se) plays an important role in suppressing oxidative stress and has huge potential in ONFH treatments. However the Se has a narrow margin between beneficial and toxic effects which make it hard for therapy use in vivo. In order to make the deficiency up, a control release of Se (Se@SiO2) were realized by nanotechnology modification. Porous Se@SiO2 nanocomposites have favorable biocompatibility and can reduced the ROS damage effectively. In vitro, the cck-8 analysis, terminal dexynucleotidyl transferase (TdT)-mediated dUTP nick end labeling (TUNEL) stain and flow cytometry analysis showed rare negative influence by porous Se@SiO2 nanocomposites but significantly protective effect against H2O2 by reducing ROS level (detected by DCFH-DA). In vivo, the biosafety of porous Se@SiO2 nanocomposites were confirmed by the serum biochemistry, the ROS level in serum were significantly reduced and the curative effect were confirmed by Micro CT scan, serum Elisa assay (inflammatory factors), Western blotting (quantitative measurement of ONFH) and HE staining. It is expected that the porous Se@SiO2 nanocomposites may prevent steroid-induced ONFH by reducing oxidative stress.

Biocompatibility test and direct flow cytometry analysis. The cartilage cells were diluted into single cell suspensions and seeded into 96-well plates (1 × 10 4 cells/well) with the culture medium. After 24 hours, the upper medium was exchanged with a medium containing different concentrations of porous Se@SiO 2 nanocomposites that were dispersed by ultrasound. After additional 24 hours, a 10% cck-8 (DOJINDO, Japan) solution was added to each well, and the plates were incubated for 1 hour in the incubator. Then, the absorbance was measured at 490 nm using a micro-plate reader. Before a comparison, the absorbances were normalized to control groups without porous Se@SiO 2 nanocomposites.
Before the flow cytometry analysis, the cartilage cells were cultured with the porous Se@SiO 2 nanocomposites at a concentration of 40 μ g/ML based on the results of the cck-8 assay for 24 hours. Then, the cartilage cells were digested and washed twice with blinding buffer (BD, USA) and analyzed directly on an Accuri C6 at the fluorescence channels for FITC, PE/PI, 7AAD and APC. For the apoptosis analysis, the cells were incubated with Scientific RepoRts | 7:43914 | DOI: 10.1038/srep43914 5 μ L of APC-Annexin V and 5 μ L of PI for 15 min at room temperature (25 °C) in the dark, and then 400 μ L of 1 × Binding Buffer was added to each tube. The cell suspension was then analyzed on an Accuri C6.

Protection by the porous Se@SiO 2 nanocomposites as assessed by suppression of ROS. H 2 O 2
was considered a classical simulation of ROS 38 and widely used [39][40][41] which was also used in this experiment.
Cartilage cells were plated into 6-well plates (1 × 10 6 cells/well) and were then pre-stimulated by the porous Se@ SiO 2 nanocomposites in a culture medium at a concentration of 0 μ g/ML, 20 μ g/ML, and 40 μ g/ML for 24 hours, respectively. Subsequently, the cells in the plates and the EP tubes (collected by digestion for the flow cytometry test) were stimulated with 50 μ M H 2 O 2 (Sigma-Aldrich St. Louis, MO, USA) in PBS (1.5 mL/well, 500 μ L/EP tube). 30 min later, DCFH-DA at its working concentration was added after 2 washes with PBS (1000 rpm, 5 min for intervention in EP tube). After 15 min of staining at 37 °C, the plates were directly observed with a fluorescence microscope (Leica, German). Cells in the EP tube were analyzed by flow cytometry.
For the cell supernatant tests, DF-12 (1.5 mL/well) was added to the 6-well plates after the H 2 O 2 stimulation. IL-1β , IL-4 and IL-6 were assessed 24 hours later by Elisa Kits (Neobioscience, China).
Protection by the porous Se@SiO 2 nanocomposites as assessed by apoptosis tests and cell activity. For the TUNEL assay, the treatment group, with or without a 24-hour pre-stimulation with 40 μ g/ML of the porous Se@SiO 2 nanocomposites, was stimulated with 200 μ M H 2 O 2 in DF-12, while the control group was stimulated with DF-12 for 24 hours. The TUNEL assay was then used to assess DNA fragmentation by a commercially available kit (In Situ Cell Death Detection Kit, fluorescein, Roche, Indianapolis, IN, USA). Briefly, the fixed cells on the slides were washed three times for 5 min with PBS and permeabilized with 0.1% (v/v) Triton X-100 containing 0.1% (w/v) sodium citrate for 2 min. The samples were then incubated in 50 μ L of TUNEL reaction mixture for 1 h at 37 °C in a dark and humidified atmosphere. Subsequently, 6-diamidino-2-phenylindole (DAPI) was used for staining of the nuclei. Positive TUNEL staining was observed under a fluorescence microscope. The sperm TUNEL index was evaluated by determining the ratio of the number of TUNEL-positive cells to that of total cells in each of the ten fields of vision.
For the flow cytometry analysis, cells received pre-stimulation or no stimulation with the porous Se@SiO 2 nanocomposites in 6-well plates for 24 hours, and then 200 μ M and 500 μ M H 2 O 2 in DF-12 were each used separately as inducements. After 24 hours, the suspension cells and adherent cells were collected and measured with an annexin V/APC apoptosis detection kit (eBioscience, USA). Briefly, the cells were trypsinized and pelleted by centrifugation, washed once with ice-cold PBS, and resuspended in 1 × Binding Buffer at a concentration of 1 × 10 6 cells/ML, from which 100 μ L of cell suspension (1 × 10 5 cells) was transferred to a 1.5 mL EP tube. Staining was then completed that is outlined above.
For the cck-8 assay, the cartilage cells were diluted into single cell suspensions and seeded into 96-well plates (1 × 10 4 cells/well) with a culture medium. After 24 hours, the upper medium of the experiment group was exchanged with a medium with 40 μ g/mL of the porous Se@SiO 2 nanocomposites, while the control group had a replacement of a medium without the porous Se@SiO 2 nanocomposites. After a 24-hour stimulation, different concentrations of H 2 O 2 were used for stimulation. After additional 24 hours, a 10% cck-8 (DOJINDO, Japan) solution was added to each well, and the plates were incubated for 1-2 hours in the incubator. Then, the absorbance was measured at 490 nm using a micro-plate reader.

Animal experiments
Animal preparation. This study was performed following the National Institutes of Health guidelines for the use of experimental animals, and all animal protocols were approved by the Institutional Animal Care and Use Committee of Shanghai Jiaotong University. Male Sprague-Dawley (SD) rats (weight 250-300 g; age of 12 months; SPF class) were obtained from the experimental animal center of Shanghai Jiao Tong University. The rats were bred and maintained under a 12/12-hour light-dark cycle with free access to food and water. The room temperature was set to 18 °C-25 °C, and the relative humidity was set to 40-60%.
Testing changes in ROS levels in serum after stimulation by the porous Se@SiO 2 nanocomposites in vivo. Prior to the experiments on the ONFH model, the porous Se@SiO 2 nanocomposites were directly injected into adult rats intraperitoneally at doses of 0 mg/kg, 1 mg/kg, 2 mg/kg and 4 mg/kg. 24 hours later, the serum from each group was cultured as outlined in a previous study 42 , and the ROS (Nanjing Jiancheng, China) was measured with an Elisa kit.
ONFH animal model establishment, treatments and sample collection. According to a previous study 43 , an early stage SANFH model was induced using a combination of lipopolysaccharides (LPS) and methylprednisolone (MPS). 36 rats were randomly divided into a control group (group A), model group (group B) and porous Se@SiO 2 nanocomposite group (group C), each consisting of 12 rats. Male SD rats from group B and group C were intravenously injected with LPS (10 μ g/kg body weight). After 24 hours, three injections of MPS (20 mg/kg body weight) were administered intramuscularly every 24 h for 5 days. To prevent infection, each rat was intramuscularly injected with 100,000 U of penicillin. Based on the toxicity study data of selenium nanoparticles in rats 37 and control release capacity of porous Se@SiO 2 nanocomposite 35 . The rats in group C were injected intraperitoneally with 1 mg/kg of the porous Se@SiO 2 nanocomposites per day for 14 days, beginning 4 weeks after the MPS administration. The model group (group A) was fed and housed under identical conditions but received saline injections.
The rats in all groups were sacrificed by an overdose of anesthesia at 8 weeks after the first MPS injection, and the femoral heads and blood samples were harvested. Blood samples from all groups were collected in containers without anticoagulant, thus allowing clot formation. The blood was centrifuged at 1,200× g for 10 minutes. The serum was stored at − 80 °C until further analysis. The left femoral heads of all rats were preserved in a − 70 °C cryogenic freezer immediately after sacrifice, and the proteins were isolated for Western blot analysis. The right femoral heads were collected and immediately fixed with 10% formalin (0.1 M phosphate buffer, pH 7.4) at 4 °C for 24 hours. Then, the samples were used for Micro CT scanning and HE staining tests following previous protocols 44 . Evaluation techniques. Micro CT procedure. According to a previous study 44    Western blot analysis. The protein expression levels of IL-1β 45 , collagen II 46 , MMP-13 47 and aggrecan 48 in the femoral head tissues obtained from rats in the different groups were detected by Western blot analysis. The Western blot protocol and semi-quantitative analysis were carried out following the protocols of a previous study 43 . Antibodies, obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA), against the following were used: IL-1β , collagen type II, MMP-13, aggrecan, and GAPDH.
Hematoxylin-eosin staining. According to a previous study 49 , the femoral heads harvested from the animals were fixed with 4% formalin and were then washed with PBS buffer. Sequentially, they were decalcified with 10% EDTA and neutralized with sodium sulfate buffer for approximately 4 weeks. After decalcification, the tissues were embedded in paraffin and cut in the coronal plane into 4-μ m thick sections with a microtome. Then, H-E staining was processed for the micro-structure observation.

Serum Biochemistry.
Serum from each groups were tested. The activities of blood serum marker enzymes, such as alanine transaminase (ALT), alkaline phosphatase (ALP), aspartate transaminase (AST), creatinine, and urea, were measured using a Roche kit (Penzberg, Germany) and analyzed spectrophotometrically using the Hitachi Analytical Instrument (Roche Diagnostic GmbH, Mannheim, Germany).

Statistical analysis.
Quantitative data are expressed as the mean ± SD. Data were analyzed using SPSS 21.0 software (IBM, Armonk, NY, USA). For comparisons of means among multiple groups, one-way ANOVAs followed by LSD tests was performed. Differences were considered statistically significant when P < 0.05.

Results
A series of tests demonstrated that the nanocomposites used in the following experiments were the porous Se@ SiO 2 nanocomposites. The phase structure of the resulting nanocomposites was examined by the XRD pattern, as shown in Fig. 1a. Several well-defined characteristic peaks, such as (100), (011), (110) and (012), exhibited the hexagonal phase, referenced by the standard Se phase (JCPDS card no. 65-1876). In addition, the XRD pattern of the Se@SiO 2 nanocomposites showed a steady increase in the low angle region, which is due to amorphous silica. It can be observed in Fig. 1b that the homogeneous nanocomposites have a diameter of about 55 nm, which many very small nanoparticles (less than 5 nm) were interspersed from the center to the surface. The

Figure 5. Two-dimensional pictures of normal bone from normal rats, and osteonecrotic bones from steroid-induced ONFH rats with or without the treatment by the Porous Se@SiO 2 nanocomposites.
The necrosis area (region of low-density) significantly reduced by Porous Se@SiO 2 nanocomposites therapy, yet not completely cured referring to control. dispersed nanoparticles have an interplanar spacing of 0.218 nm, matching the spacing for the (110) crystal planes of the standard hexagonal Se (Fig. 1b inset), which further confirmed that the small nanoparticles dispersed in the silica were Se nanocrystals. After treatment with hot water, the porous Se@SiO 2 nanocomposites were formed (Fig. 1c,d). Our previous characterization of BET showed that the Se@SiO 2 nanocomposites are porous 35 . The PVP (pyrrolidinovalerophenone) had permeated into the silica shells, and the channels were distributed in the Se@SiO 2 nanocomposites, leading to the slow release of very small nanoparticles from the porous Se@SiO 2 nanocomposites. Meanwhile, the release capacity were re-confirmed (Fig. S1).
Cells were identified as cartilage cells by staining with aggrecan and collagenase type I (Fig. 2a,b). Co-culturing with the porous Se@SiO 2 nanocomposites was demonstrated to have no significant influence on the flow cytometry (Figs S2 and S3) and proved to be safe by flow cytometry at a concentration of 40 μ g/ML (Figs 2c, S4). The porous Se@SiO 2 nanocomposites showed no significant cytostatic action under a concentration of 40 μ g/ML compared with the control group (Fig. 2d). The fluorescence microscopic observation indicated that pre-stimulation with the porous Se@SiO 2 nanocomposites might significantly decrease the expression of the ROS (Fig. 3) after H 2 O 2 exposure, which was re-confirmed and quantitatively analyzed by flow cytometry analysis (Fig. S5). Pre-stimulation with the porous Se@SiO 2 nanocomposites could significantly reduce the expression of IL-1β , IL-4 and IL-6 in a concentration-dependent manner (Fig. S6, P < 0.05).
After 24 hours of stimulation with 200 μ M H 2 O 2 , the cells pre-stimulated with 40 μ g/ML of the porous Se@ SiO 2 nanocomposites demonstrated less damage than the control groups. The apoptosis analysis showed that the porous Se@SiO 2 nanocomposites significantly decreased the apoptosis rates caused by H 2 O 2 (Fig. 4a,b, *P < 0.05), which was re-confirmed by the flow cytometry analysis (Fig. S7, *P < 0.05) and the cck-8 assay (Fig. 4c, *P < 0.05) at different H 2 O 2 concentrations.
After 24 hours through the intraperitoneal injection of the porous Se@SiO 2 nanocomposites, the ROS levels in the serum were significantly decreased (Fig. S8, *P < 0.05 in comparison with the other groups). 8 weeks after the steroid injection, the weight of the rats was maintained at 426 ± 32 g, with no significant differences detected among the groups.
The porous Se@SiO 2 nanocomposite treatment was performed for 4 weeks after the steroid exposure to avoid the possible influence of the model inducement. Under this situation, the protective role of the porous Se@SiO 2 nanocomposites in the early stage of ONFH were evaluated.
A CT scan indicated the therapeutic effects of the Porous Se@SiO 2 nanocomposite treatment (Fig. 5), and the data analyses on the levels of the microstructural parameters showed significant differences (Fig. S9).
Serum ALT, ALP, AST, creatinine, and urea in each group had no significant differences (Table S1). IL-1β and MMP-13 were classical damage indexes in cartilage reaches that can objectively evaluate the necrosis situation. While collagen type II and aggrecan were both constructional and functional protein in cartilage. The changes of these four protein assessed by Western blot analysis can fully reflect the specific situation. Western blot analysis for both subchondral bone and cartilage showed that the expression of IL-1β , collagen type II, MMP-13 and aggrecan increased in the model group. After the Porous Se@SiO 2 nanocomposite injection, expression significantly decreased (Figs 6 and 7, P < 0.05), except for collagen II expression in cartilage (Fig. 6). The blot is representative of subchondral bone, normalized to GAPDH, and statistical analysis was performed (n = 12, *P < 0.05) (b, d); the blot is representative of cartilage, normalized to GAPDH, and statistical analysis was performed (n = 12, *P < 0.05).
Hematoxylin-eosin staining were used for structural observation of the femur head. Serious destruction of the femur head was induced for 8 weeks after methylprednisolone administration. After the Porous Se@SiO 2 nanocomposite injection, fewer necrotic areas (cavity beneath the cartilage surface) were detected (Fig. 8).
Still, the mechanism of ONFH are still unclear, but ROS and ROS suppressing treatments indicated the close relationship oxidative stress have to ONFH. A newly reported review of hyperbaric oxygen therapy in the treatment of osteonecrosis of the femoral head confirmed the therapeutic function antioxidants have in ONFH 50 . Thus by decreasing the ROS levels, other elements should also have the capacity to retard the progress of ONFH. However, antioxidants, such as coenzyme Q10 (CoQ10), may lead to nausea, upset stomach or loss of appetite. In addition, hydrogen-rich saline is not stable enough, and stem cell factor (SCF) and grape seed proanthocyanidin are costly as standard treatments. Compare to which, thanks to the SiO 2 -coated structure, significant advantages Porous Se@SiO 2 nanocomposites have in both economic and stability (room temperature store) are convinced.
Moreover, porous structure gave Se@SiO 2 nanocomposites potentials to be multifunctional (combined with drugs) and made them slow-released. Controlled-release systems play special roles in disease treatment 51,52 . Compared to normal Se nanoparticles, Se in the Se@SiO 2 nanocomposites is limited by SiO 2 . Accompanying the entrance of PVP into an aqueous solution, trace Se can be released into the solution 35 . By controlled-release, the porous Se@SiO 2 nanocomposites may have advantages in biosafety (Fig. 2d) and in vivo stability. Because the process of ONFH is long and progressive 53 , the controlled-release capacity benefits the sustaining of the ROS inhibition. So, SiO 2 -coated ultrasmall Se particles may help to delay the onset or reduce the serious outcome of the ONFH.
The advantages and characteristics porous Se@SiO 2 nanocomposites have made them an ideal therapy to ONFH. The possible mechanism may be the ROS suppressing. Nano-materials that can mediate the ROS expression are not unique; some function by direct contact, and some function by anti-bacterial properties 54 , while others help to maintain structural stability and improve bio-safety 55 , even help to induce the cell apoptosis program via the ROS 56 . However, the porous Se@SiO 2 nanocomposites may not only reduce the expression of the ROS directly but also provide essential elements that help to comprise the intracellular pool against oxidative stress 57,58 . Accumulating evidence supports the idea that Se nanoparticles have antioxidant effects 59 . These effects have been shown to increase the activities of both GPX and glutathione S-transferase and induce less oxidative stress 34,59 . The same consequences were observed in this study, in which the porous Se@SiO 2 nanocomposite simulation significantly decreased the ROS levels and improved the cells' tolerance to H 2 O 2 (Fig. 4), with the ROS levels in serum demonstrating the same consequence in vivo (Fig. S7). It also has been reported that by inhibiting the activation of the PI3K/AKT and ERK signaling pathways and endoplasmic reticulum stress, Se can suppress oxidative-stress-enhanced vascular smooth muscle cell calcification 60 , further reducing the levels of IL-1β , TNF-α , oxidative stress, and NF-κ B activation 61 . On the contrary, a diet with a Se deficiency weakens antioxidant capacity 62 . However, despite the acute side effects associated with toxicity 63 , pre-stimulation with porous Se@SiO 2 nanocomposites did help to reduce the expression of IL-1β , IL-4, and IL-6 after exposure to H 2 O 2 in a concentration-dependent manner (Fig. S5). This change matched the decreased IL-1β level in the femoral head (Figs 6 and 7).
This improvement was reflected not only by Micro CT (Fig. 5) and the histological images ( Fig. 8) but also by the protein levels of tissue biopsies (Figs 6 and 7). In the CT imaging, a low attenuation section in the femoral head refers to collapse and necrosis. Observed in 3-dimensional images, after treatment by the porous Se@SiO 2 nanocomposites, the percentage of low attenuation area was reduced. By a quantitative comparison (Fig. S8), significant differences were detected among these groups confirmed the curative effects. Meanwhile, the tissue biopsy directly showed the improvement in the femur head structure in treatment group (Fig. 8). Thus, it is receivable that after porous Se@SiO 2 nanocomposite treatment, anatomic integrity of femoral head was maintained.
Western blot investigation showed a significant increase in IL-1β and MMP-13, which indicates that damage had occurred in the cartilage and subchondral bone, which decreased after the porous Se@SiO 2 nanocomposite treatment. The expression of aggrecan and collagen type II indicated a compositional variation after steroid inducement, as their variation combined the structural and functional changes in the femoral head, which were also significantly recovered after the porous Se@SiO 2 nanocomposite treatment. These results confirmed the local phenomenon and possible mechanism that ONFH may started with IL-1β and MMP-13 damages, influenced the structural and functional protein aggrecan and collagen type II and finally leads to necrosis (Fig. 6). By suppressing IL-1β and MMP-13, porous Se@SiO 2 nanocomposites can significantly reduce the damage of ONFH.
However, in this experiment with the porous Se@SiO 2 nanocomposites injection, only its possible function in the early stage of ONFH was tested, and the potential preventive effects of the porous Se@SiO 2 nanocomposites, according to the basic function of ROS inhibition, were not demonstrated. For preventive treatment, it has to be proved that the porous Se@SiO 2 nanocomposites have no influence on the therapeutic effects of the steroids.
Another study system will be needed and will be reported in the future.

Conclusions
In summary, the porous Se@SiO 2 nanocomposites could reduce the ROS levels, protect cells from H 2 O 2-induced apoptosis and reduce inflammation by inhibiting oxidative stress. Same effects were observed in vivo and eventually demonstrated the significant protective effect in disease management.
The porous Se@SiO 2 nanocomposites have advantages of stability, economy, and enormous multifunctional potentials compared to other methods. Therefore, make porous Se@SiO 2 nanocomposites an ideal way to protect the femoral head from osteonecrosis after steroid inducement.