Data collection and preprocessing

We investigated the development of DMT1 in osteoporotic and non-osteoporotic bone marrow mesenchymal stem cell samples by analyzing human osteoporotic and non-osteoporotic bone marrow mesenchymal stem cell samples in an online database. Data from GSE147287 were downloaded from the Gene Expression Omnibus database. This database included single cell sequencing of osteoporosis (GSE147287) and osteoarthritis tissues (GSM4423511). Further down-clustering analysis, selection of differentially expressed genes, differential analysis, and acquisition of marker genes were performed using R software (version 4.2.0: R Foundation for Statistical Computing, Vienna, Austria) and its Seurat package (version 4.0) single cell data analysis software set developed by Satija Laboratories (New York, USA). Using this method, DMT1 was found to have potential in the development of osteoporosis, and pseudotime analysis revealed up-regulation of this gene expression in cells that developed OP.

Pseudo-time analysis

Single cell pseudo-time analysis using Monocle2 (http://cole-trapnelllab.github.io/monocle-release) with DDR-Tree and its default parameter was used. Before monocle analysis, marker genes of Seurat clustering results and raw expression counts of the cells that passed filtering were selected. Based on pseudo-time analysis, branch expression analysis modeling was used for branch fate-determined gene analysis.

Cell culture

The MC3T3-E1 cell line derived from mice (iCell, Shanghai, China) was cultured in α-MEM (Wisent, China) supplemented with 10% fetal bovine serum (FBS) and incubated at 37℃ in a 5% CO2 incubator. Before seeding, MC3T3-E1 cells were cultured and transferred using MEM medium supplemented with 10% FBS (Gibco, Grand Island, NY, USA) and 100 µ/mL penicillin-streptomycin-amphotericin B (Biotech, China), with daily fresh medium replacements.

Cell viability assay

A cell viability assay was used to determine the cytotoxicity of MT (Sigma Aldrich, Shanghai, China) using the CCK-8 method. MC3T3-E1 cells were placed in 96-well plates at a density of 1 × 104/well. They were then treated with varying amounts of MT (0, 5, 10, 25, 50, and 100 µmol/L) for 3 days. Then, 10 µL of CCK-8 cell viability buffer was added to each well, followed by incubation at 37℃ for an additional 1 h. Finally, the absorbance at 450 nm was determined using a microplate reader.

Preparation of PAP, chitosan, and chitosan@PAP

Materials-chitosan was from Guangzhou Yunmei Technology (Guangzhou, China). The amino groups of p-phenylenediamine and N-phenyl-1,4-phenylenediamine, which was shielded with butane diacid anhydride, were dissolved in a solution containing dimethylformamide and HCl. Toluene (10 mM) was used to azeotropically distill the hydroxyl-capped PLA. Next, a flame-dried glass reactor was filled with 2 mmol of purified PLA, 1 mmol of emeraldine carboxyl-capped aniline pentamer (EMAP), 5 mmol of DCC, 5 mmol of DMAP, and 15 mL of NMP. Following the reaction, filtration was used to eliminate dicyclohexylurea. The copolymer present in the filtrate was precipitated using ethanol and subsequently dissolved in CHCl3. PAP was then dried at room temperature under vacuum. To prepare the chitosan Complex Coated PAP granules, ethyl acetate was initially used to disperse PAP (10 g). Water was then added to agglomerate the dispersed particle mixtures, which were then agitated. After the separation process, the resulting particles were dried. The granules that were screened were dispersed in the chitosan solution and subsequently coated with chitosan. Following 30 min of stirring, the granules that had been coated were isolated, rinsed with water, and subsequently dried in a desiccator. Sieve analysis was used to determine the size distribution of the desiccated granules.

Cytotoxicity assessments of chitosan, PAP, and chitosan@PAP

Cell viability was assessed using the CCK-8 assay (Biosharp, Hefei, China) as per the instructions provided by the manufacturer to confirm the cytotoxic effects of chitosan, PAP, and chitosan@PAP. Briefly, the MC3T3-E1 cells were placed in a 96-well dish at a concentration of 3 × 104 cells/mL and left to incubate for a period of 1 d. The MC3T3E1 cells were then exposed to chitosan, PAP, or chitosan@PAP in a culture medium for a duration of 3 d. Subsequently, 10 µL of CCK-8 was added to the culture and incubated at 37 °C in the absence of light for 3 h. A microplate reader was then used to measure the absorbance at 450 nm.

Quantitative RT-PCR

MC3T3-E1 cells were cultured in 6-well plates with medium at a density of 105 cells per well. The cells were subjected to treatment with MT at concentrations of 0, 5, 10, 25, 50, and 100 µmol/L. After 3 d, the RNA from cells was obtained using a rapid extraction kit designed specifically for RNA. Subsequently, the RNA was converted into cDNA using a reverse transcription kit (Beyotime, Shanghai, China). Next, SYBR Green PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) was used to conduct real-time quantitative PCR. To ensure data accuracy, 40 cycles of PCR were performed with the following conditions: 10 min at 94℃, 15 s at 95℃, and 60 s at 60℃. In addition, six replicate wells were used for all reactions.

In vitro release of chitosan@MT and chitosan@MT/PAP

After the gel formed, 100 µL of a liquid solution with a concentration of 50 μm/L of MT were added to chitosan/chitosan@PAP and stirred, shaked for 2 h, then placed in a 24-well plate with 500 µL of phosphate-buffered saline (PBS). The well plates were placed in a shaking incubator at 100 rpm and incubated for 14 d at 37 °C. The supernatant was collected with fresh PBS (0.1 M) at specific intervals [0 day, 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, 7 days (1 w), 10 days, and 14 days (2 w)]. The supernatants were then preserved at -80 °C until ready for the ELISA test. The concentration of MT in each sample was evaluated using a Human MT ELISA Kit (Fine Test, China) following the manufacturer’s instructions, at a wavelength of 450 nm. The measurement was conducted six times for each time period, and the total quantity of MT was computed and graphed over time.

Microparticle characterization

The morphology of microparticles was determined using scanning electron microscopy (SEM; Gemini 2, Zeiss, Oberkochen, Germany). To prepare the samples, a small amount of the microparticle mixture was placed on the surface and left to dry for an entire night. The microparticles were then analyzed using INVENIO-R spectrometer Fourier transform infrared spectroscopy (FTIR; Bruker, Billerica, MA, USA). The dried hydrogels were scanned using a Micro CT (PE Quantum GX2; Perkin Elmer, Waltham MA, USA) and the xx software was used to calculate the percentage of total porosity.

Immunohistochemistry analysis

Immunohistochemistry was used for detection of RUNX-2 proteins. The MC3T3-E1 cells were cultured in a 6-well plate at a density of 104 cells per well. They were then divided into four groups: Control, chitosan, chitosan@PAP, chitosan@MT, and chitosan@MT/PAP to co-cultured. Following a 1-week co-culture period, MC3T3-E1 cells were placed on medium at a concentration of 5 × 105 cells/mL. Next, the medium was supplemented with osteogenic induction medium and incubated for a duration of 48 h. Following fixation of cells in 4% paraformaldehyde for 30 min, the slides were treated with goat serum to prevent nonspecific binding. Subsequently, they were incubated with primary antibody (antiRUNX-2, 1:200; Abcam, Shanghai, China) overnight at 4ºC. Next, the medium was treated with RUNX-2 conjugated secondary antibody (at a dilution ratio of 1:50; source and address) for 30 min at ambient temperature. Then, images were captured using a microscope.

Immunofluorescence staining

The MC3T3-E1 cells were cultured in a 6-well plate at a density of 104 cells per well. They were then divided into five groups: Control, chitosan, chitosan@PAP, chitosan@MT, and chitosan@MT/PAP for co-culture. Following 1 week, the MC3T3-E1 cells were rinsed using PBS, then treated with 4% paraformaldehyde for 30 min, and subsequently permeabilized with Triton X100 for 10 min. Next, the cells were cultured with primary anti-RUNX-2 antibody at a temperature of 4℃ for the entire night. Afterward, they were stained with anti-RUNX2 and Alexa Fluor 488 Phalloidin (Molecular Probes, Eugene, OR, USA) in the absence of light for 1 h. Following a 10 min fixation period using mounting fluid containing 4’,6-diamidino-2-phenylindole, the images of cells were captured using the EVOS M5000 cell imaging system (Thermo Fisher Scientific).

Establishing an osteoporosis mouse model and in vivo experiments

A total of 32 C57BL/6 female mice, 8-weeks-old and weighing approximately 20 g each, were purchased from Ruige (Shenzhen, China). These mice were then randomly assigned to four groups: Control group, chitosan@PAP group, MT group, and chitosan@MT/PAP group. After anesthesia, all mice underwent surgery, and the incisions were closed following the removal of both ovaries and the surrounding adipose tissues. After a period of 4 weeks, the chitosan@PAP/MT group received a single injection of 10 mg/kg chitosan@MT/PAP. The chitosan@PAP group was injected with the same dose as the chitosan@MT/PAP group. Similarly, the MT group received the same dose of MT, and the control group received the same dose of saline. Following a 4-week intervention, mice were sacrificed and the femur was extracted and immobilized in a 4% paraformaldehyde solution. The mouse femur in PBS at room temperature were used to observe the bone microstructure. The femur bones underwent hematoxylin & eosin (HE) staining and micro-computed tomography (CT) analyses. Micro-CT was utilized for quantitative determination in both two-dimensional (2D) and three-dimensional (3D) formats, following previously established protocols [27,28,29]. The Siemens Preclinical Imaging System (Siemens, Amsterdam, The Netherlands) was used to measure microstructure parameters of cancellous bone, including trabecular number (Tb.N), trabecular thickness (Tb.Th), bone mineral density (BMD), and bone volume/total volume (BV/TV). These measurements were conducted using the multislice mode in standard resolution.

Statistical analysis

The mean ± standard deviation of quantitative data obtained from a minimum of six biological replicates is presented in this study. Statistical significance was evaluated using SPSS 20.0 statistical software for Windows (SPSS, Chicago, IL, USA). Student’s t-test and analysis of variance were used to determine the distinctions between 2 and > 2 groups. Groups were considered significantly different at a value of P < 0.05.