Fabrication and characterization of materials

The primary scaffolds were synthesized by a straightforward one-step approach as previously described49. Briefly, 8.0 g of P123 was placed into a mixture of HNO3 (120 mL, 2 mol·L−1) and distilled water (30 mL, 2 mol·L−1) and stirred until dissolved. Then, SrCl·6H2O (0 g, 0.67 g, or 1.34 g to form three distinct mixtures), 13.4 tetraethoxysilane, 2.8 g Ca(NO3)·4H2O, and 1.46 g triethyl phosphate were added while stirring at room temperature. The three mixtures were transferred into an autoclave and heated at 700 °C for 5 h. Next, scaffolds were fabricated by a simple powder pressing process, and polyvinylpyrrolidone and polyethylene glycol were used as adhesives and pore-forming agents. Subsequently, 3-aminopropyltrimethoxysilane (97%) was employed to graft the amino-functional groups onto the fabricated scaffolds. The three groups of fabricated scaffold samples made from the three aforementioned mixtures were denoted as N-MBG, 1Sr-N-MBG, and 2Sr-N-MBG, respectively50. SEM (Hitachi S4800 electron microscope, Tokyo, Japan) was employed to investigate their surface morphology, and inductively coupled plasma–mass spectrometry (NuInstruments, Wrexham, UK) was used to detect the released concentration of Sr ions by collecting the extract liquid of different scaffolds on Days 1, 3, 5, 7, and 14.

Preparation of osteoporotic model and culture of OVX BMSCs

Experimental protocols were approved by the Animal Care and Experiment Committee, Ninth People’s Hospital51, and all operations were carried out under sterile conditions. An osteoporotic model was induced by OVX in female Sprague–Dawley rats. Briefly, the enterocoel was exposed by a minimally bilateral lumbar incision, and then ovaries were pulled out and removed once vessel ligation was finished. Antibiotics were administered postoperatively. Three months after ovariectomy, the animals were sacrificed, the femora were harvested, and both ends were cut off. The bone marrow from the exposed medullary canal was flushed out and centrifuged to obtain a pellet of primary OVX BMSCs, which was suspended, placed in a dish, and incubated. The culture medium was replaced every 2–3 days to discard nonadherent cells under culture conditions (5% CO2 and 37 °C). Cells in passages 2–4 were used for the following experiments52.

Biocompatibility investigation

OVX BMSCs were seeded onto different samples at a density of 2.0 × 104. A live/dead assay was carried out to determine cellular fate by a Calcein/PI Cell Viability/Cytotoxicity Assay Kit after culturing for three days. According to the manufacturer’s instructions, the working solution was prepared and added to different samples (200 μL per well). After incubation for 30 min, Calcein AM was used to stain living cells, which showed green fluorescence, while red fluorescence indicated dead cells labeled by propidium iodide (PI). The relative live/dead fluorescence intensity was calculated, and statistical analysis was performed. The cell morphology was also observed by SEM. Each sample was washed with phosphate-buffered saline, fixed in 3% glutaraldehyde, dehydrated in increasing concentrations of ethanol, and then sputter-coated with platinum for final observations. The proliferation capability was detected by Cell Counting Kit-8 (CCK-8) analysis. On Days 3 and 7, CCK-8 (200 μL per well) was added to DMEM and incubated for 1 h. Afterward, the absorbance value was measured at 490 nm by an ELX ultra microplate reader (BioTek, USA). All operations were performed in triplicate.

Osteogenic ability evaluation in vitro

For the alkaline phosphatase (ALP) activity quantitative assay, cells seeded on samples were collected and incubated with p-nitrophenyl phosphate (pNPP, Sigma, St. Louis, MO, USA) at 37 °C for 30 min. The results were measured and displayed as optical density (OD) values at 405 nm. The Bradford method was used to detect the total protein content. All operations were performed in triplicate. Real-time PCR assays were also carried out. Briefly, the RNA of cells was extracted after 14 days of culturing by TRIzol (Invitrogen, USA). The harvested RNA was used to synthesize DNA with a cDNA synthesis kit (Takara, Japan). By the reverse transcription-polymerase chain reaction system (Bio-Rad, USA), the expression levels of osteogenic and angiogenic markers (OCN, Runx2, and VEGF) were detected. The primer sequences are shown in Table S1. All expression levels were normalized to β-actin, and the operations were performed in triplicate. Moreover, to investigate the biological performance of the Sr ions released from the scaffolds, cells were incubated in the extracted liquid of each material for 14 days, and ARS staining was used to investigate extracellular matrix mineralization. Briefly, cells were fixed in 4% paraformaldehyde, and a 0.1% ARS working solution was prepared according to the manufacturer’s instructions and used for the following study.

Angiogenetic capability ability evaluation in vitro

Tube formation analysis was conducted for HUVECs cultured in the extract of Sr-N-MBG scaffolds to determine their angiogenic ability in vitro. All the experimental instruments, including culture plates and tips, were precooled at −20 °C before the plates were coated with Matrigel (Corning, USA) at 150 μL per well. The plates were incubated for 30 min at 37 °C to solidify the Matrigel. Then, HUVECs (ScienCell, USA) were suspended and seeded onto the preplated Matrigel at 2 × 105 cells per well. After the cells were attached, the scaffold material extracts were added to the medium, and microscopy was used to detect tube formation at 4 h. Furthermore, western blotting was carried out to detect the VEGF expression levels in the HUVECs. Briefly, HUVECs were cultured in Sr-N-MBG scaffold extract for seven days. The cells were lysed with the protein extraction reagent, which contained a protease inhibitor cocktail, phosphatase inhibitor cocktail, and phenylmethanesulfonylfluoride (PMSF) (Kangchen, China), and the concentration of acquired protein was calculated by a Bio-Rad protein assay kit. Equivalent amounts of protein were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and electrotransferred to polyvinylidene difluoride (PVDF) membranes (Pall, USA). The PVDF membranes were incubated with specific primary VEGF antibodies (CST, USA) at a dilution of 1:1 000 overnight at 4 °C. Finally, the PVDF membranes were visualized by horseradish peroxidase (HRP)-conjugated secondary antibodies (Beyotime, China) by ECL Plus reagents (Amersham Pharmacia Biotech, USA) using a UVItec ALLIANCE 4.7 gel imaging system.

Surgical procedure for ectopic osteogenesis in nude mice

Six nude mice were selected and anesthetized for implantation. In each mouse, a subcutaneous dorsal pocket was created by blunt dissection, three groups of scaffolds seeded with OVX BMSCs were implanted, and the wounds were sutured. Four weeks later, the mice were sacrificed, and the scaffolds together with 1-cm soft tissue were extracted and fixed. After decalcification, they were embedded in paraffin and subsequently fabricated into 4-μm-thick slices by a microtome (Leica, Germany). The sliced sections were subjected to HE staining to investigate bone formation.

Animal surgical procedure and histological investigation

Critically sized defects of 5 mm in diameter were created in calvarial bones in the osteoporotic model. Subsequently, each scaffold (5 mm in diameter) was used to implant the defects, and antibiotics were administered postsurgically53. After eight weeks, the animals were sacrificed, and the skulls of the surgery regions were harvested. The bone regeneration outcome was detected by micro-CT (SkyScan 1176, Belgium) and 3-D Creator software (Scanco Medical, Switzerland). Moreover, as previously described, corresponding parameters (BMD, BV/TV, Tb. Th, and Tb. N) were also calculated to obtain quantitative results. The skull specimens were prepared into halves along the sagittal plane for subsequent histological observation. After decalcification, they were embedded in paraffin and subsequently trimmed into 4-μm-thick slices. The slice sections were subjected to HE staining and immunohistochemical (IHC) staining for CD31 and VEGF (Abcam) to investigate bone formation and angiogenesis. The quantitative parameters were measured using Image-Pro Plus™ (Media Cybernetics, Silver Springs, MD, USA).

RNA-Seq analysis

OVX BMSCs were seeded onto the N-MBG and 2Sr-N-MBG scaffolds and incubated, and total RNA was isolated by TRIzol (Invitrogen, USA) according to the manufacturer’s instructions. The qualities of the RNA samples were assessed, and those with a 28S/18S RNA ratio ≥0.7 were acquired54. For sequence and primary analysis, Cutadapt software was employed to remove the reads that contained adapter contamination, and then HISAT2 software was used to map the reads to the genome. Then, all transcriptomes from each sample were merged to reconstruct a comprehensive transcriptome. The differentially expressed mRNAs were selected with a fold change >2 or fold change <0.5 and P < 0.05. GO enrichment and KEGG enrichment analyses of the differentially expressed mRNAs were carried out.

Detection of intracellular and mitochondrial ROS and the expression of cAMP/p-PKA

The level of intracellular ROS production was determined by labeling the OVX BMSCs seeded on the N-MBG and Sr-N-MBG scaffolds with 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) (10 × 10−6 m, Thermo Scientific) for 30 min. Moreover, to investigate the levels of mitochondrial ROS, the cells were counterstained with the selective fluorescent dye MitoSOX Red (2.5 × 10−6 m, Thermo Scientific), which targets superoxide production in the mitochondria of living cells. MitoTracker Green was employed to label the mitochondria (150 × 10−9 m, Thermo Scientific) via incubation for 30 min at 37 °C. According to the KEGG enrichment analysis results, western blotting was used to detect cAMP/p-PKA expression. Briefly, OVX BMSCs were cocultured in extracts of different materials for seven days. The cells were lysed with protein extraction reagent, and the concentration of acquired protein was calculated. As mentioned above, equivalent amounts of protein were separated by SDS–PAGE and electrotransferred to PVDF membranes incubated overnight at 4 °C with specific primary antibodies against cAMP and p-PKA (CST, USA) at a dilution of 1:1 000. Finally, they were visualized by HRP-conjugated secondary antibodies (Beyotime, China) by ECL Plus reagents (Amersham Pharmacia Biotech, USA) using the UVItec ALLIANCE 4.7 gel imaging system.

Statistical analysis

SPSS software (SPSS Inc., USA) was used, and the statistical significance was determined by analysis of variance and t tests. The symbol ⋆ indicates P < 0.05 compared with the N-MBG group, and the symbol # indicates P < 0.05 compared with the 1Sr-N-MBG group.