All animal care and experimental protocols were approved by the Institutional Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology (Wuhan, China, S780), which were according to the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals.

All male C57BL/6J mice (aged 8 weeks or 78 weeks) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. All these mice were raised in the specific pathogen-free (SPF) animal center of Tongji Medical College, Huazhong University of Science and Technology (Wuhan, China). The mice were maintained at a room temperature of 23 ± 1°C and 55 ± 5% humidity under a 12 h light/dark cycle and supplied with sterilized food and water. The irradiated corncob bedding was changed every week.

Sprague–Dawley rats (aged 3 weeks, male, weighing approximately 60–80 g) were used for collecting and culturing the BMSCs for the experiments in vitro.

The bone marrow of femurs and tibias was drawn out by flushing with a complete medium. The bone marrow cells of each rat were seeded in two 25-cm² tissue culture flasks containing DMEM/F-12 medium (HyClone, Logan, UT, United States), 10% fetal bovine serum (FBS) (Gibco, United States), and 1% penicillin-streptomycin solution (Beyotime, Shanghai, China). The BMSCs were maintained at 37°C in a 5% CO₂ humidified incubator. By changing the medium every 2 days, non-adherent cells were removed and adherent cells were cultured until confluent to 80%.

The cells were cultured on the slides (round, thickness: 0.2 mm, and diameter: 15 mm). Cell proliferation was analyzed using the BeyoClick™ EdU Cell Proliferation Kit (Beyotime, Shanghai, China) with Alexa Fluor 488, according to the protocol. Every well was treated with 500 µl of medium containing 10 µM EdU. After incubation at 37°C, with 5% CO₂ for 2 h, the cells were fixed with 4% paraformaldehyde at room temperature for 15 min and incubated with 0.3% Triton X-100 in PBS for 10–15 min. The cell nuclei were stained with Hoechst 33342. Images of three randomly selected areas of each group were taken using a laser scanning confocal microscope (A1si+/A1Rsi+, Nikon, Japan). The percentage of proliferative cells was calculated by the number of EdU-positive cells compared with all cells in different fields as already shot.

Cell apoptosis was evaluated by using the One-Step TUNEL Apoptosis Assay Kit (Beyotime, Shanghai, China). The apoptotic cells were detected using a fluorescence microscope (Olympus, Japan). Images of three randomly selected areas of each group were taken.

The senescence β-galactosidase staining kit was used to test the activity of senescence-associated β-galactosidase (SA-β-Gal) (Beyotime, Shanghai, China). According to the protocol, the senescent cells will be stained.

To induce osteogenic differentiation, the BMSCs were cultured in a commercially available osteogenic medium (Cyagen, Soochow, China) for 14 days. Thereafter, the cells were washed twice with phosphate-buffered saline (PBS), fixed in 4% paraformaldehyde for 15 min at room temperature, and stained with alizarin red (Servicebio, Wuhan, China) for 30 min. After rinsing the cells 2–3 times with PBS, the calcified nodules were visualized under an inverted microscope (Olympus, Tokyo, Japan). Similarly, the BMSCs were cultured in a commercially available adipogenic medium (Cyagen, Soochow, China) for 2 weeks. Subsequently, the cells were washed twice with PBS, fixed in 4% paraformaldehyde for 15 min at room temperature, then washed with PBS, and stained with 2% oil red O (Servicebio, Wuhan, China) for 1 h at room temperature. The BMSCs were washed again with PBS. The intracellular lipid droplets were visualized under a microscope (Olympus, Tokyo, Japan).

Total RNA was extracted from cells using the RNA isolater Total RNA Extraction Reagent (Vazyme, Jiangsu, China), following the manufacturer’s protocol. Complementary DNA was generated using the HiScript^® II Q RT SuperMix for qPCR (+ gDNA wiper) (Vazyme, Jiangsu, China). The qRT-PCR was performed in a 96-well plate with ChamQ SYBR qPCR Master Mix (Vazyme, Jiangsu, China), and it was performed on an ABI StepOnePlus real-time PCR instrument (Applied Biosystems). The mRNA levels of the target genes were normalized against that of GAPDH, which served as an internal control.

The cells were lysed by RIPA buffer supplemented with a protease inhibitor cocktail. An equal amount of proteins (containing 30 mg protein lysate) were loaded into 12% polyacrylamide gel for SDS-PAGE electrophoresis and transferred to 0.45 µm or 0.22 µm polyvinylidene fluoride membranes, which were subsequently incubated with 5% skim milk for 1 h and then probed with rabbit antibodies against Perilipin 1 (1:2000, ab3526, Abcam), FABP4 (1:500, 15872-1-AP, Proteintech), SOX2 (1:500, A0561, ABclonal), and GAPDH (1:10000, 10494-1-AP, Proteintech). The enhanced chemiluminescence substrate reagent (Millipore, Billerica, MA, United States) was used to detect immunoreactive protein bands. The gray values of the blots were analyzed by ImageJ software. The value of each blot was normalized to the value of GAPDH.

The RayBiotech Quantibody^® Array, multiplexed sandwich ELISA-based quantitative array platform, enables us to accurately measure the concentration of multiple cytokines simultaneously. In detail, the serum samples and standards were incubated on the array overnight at 4°C. The slides were washed with a proprietary buffer and then incubated with a detection antibody at room temperature for 2 h. The slides were then washed and incubated with the streptavidin-conjugated Cy3 equivalent dye at room temperature for 1 h. Thereafter, the slides were washed a final time and then dried thoroughly before scanning with a laser scanner using the Cy3 excitation profile. Then, we calculated and analyzed the cytokine concentration.

Supernatants were collected from the BMSCs. The concentrations of IL-6 and IL-10 in the supernatant were determined by using commercial ELISA kits (NEOBIOSCIENCE, China), according to the manufacturer’s instructions. The absorbance was measured by using a microplate reader at 450 nm.

The femur specimens were fixed in 4% paraformaldehyde for 24 h, decalcified with 10% EDTA in PBS for 3 weeks, embedded in paraffin, and sectioned to a 5-µm thick specimen. Sections were stained by hematoxylin and eosin (H&E) staining. For immunohistochemical staining, the sections were placed in sodium citrate buffer for 20 min at 90ºC for antigen retrieval and incubated with monoclonal anti-mouse IL-6 and IL-10 antibodies overnight at 4°C.

All the data were displayed as mean ± standard deviation (SD). ImageJ software was used for quantitative protein analysis. Data were evaluated by GraphPad Prism 7.00 software using Student’s t-test or one-way ANOVA followed by Tukey’s post hoc test. A p-value < 0.05 was considered significant.

Clinically, older patients have a higher incidence of fractures than younger patients. Hematoxylin–eosin (HE) staining of femur slices was used to assess the bone mass and fat content in young (8-week-old) mice and aged (78-week-old) mice. The bone mass in the aged group was significantly lower than that in the young group. Conversely, the fat content in the aged group was higher than that in the young group (Figures 1A,C,D). The activity of β-galactosidase was enhanced in the bone marrow cavity in the aged group, which indicated an increased percentage of the senescent cells in aged mice (Figure 1B). After long-term passage in vitro, the BMSCs in the aging group (P10) showed changes in morphology, including larger and flatter cell features (Figure 1E), and enhanced β-galactosidase activity (Figure 1F). These results indicated that the number of senescent cells increases with aging.

Stem cell senescence is caused by a combination of intrinsic and irreversible changes through circulating effectors or factors secreted by local stem cell niches (Severino et al., 2013). To determine whether extracellular factors produced by the senescent cells can alter the physiological characteristics of stem cells, proliferation and the differentiation capacity of young MSCs cultured with conditioned medium (CM) from the P3 and P10 BMSCs were evaluated. When P3 and P10 BMSC cultures reached an 80% confluency, the serum-free base medium was added after discarding the original medium, and the supernatant (CM) was collected and stored at −80°C for 48 h. As expected, the positive rate of EdU staining in the CM-P10 cultured cells was lower than that in the CM-P3 cultured cells (n = 3 and p < 0.01) (Figure 2A). The proliferation rate of BMSCs was gradually decreased when the cells were cultured in CM-P10 compared with those cultured in CM-P3 (Figure 2B). Alizarin red staining revealed that there was an evident decrease in osteoblast differentiation in the mBMSCs cultured in CM-P10 compared with those cultured in CM-P3 (Figure 2C, Supplementary Figure S1). Furthermore, the RT–PCR analysis indicated that the expressions of osteogenic markers, such as OCN, ALP, Runx2, and OSX, were decreased when the cells were cultured in CM-P10 (n = 3 and p < 0.05) (Figure 2D). Oil red O staining indicated that the number of lipid droplets in the CM-P10 group was significantly reduced compared to that in the CM-P3 group (Figure 2E, Supplementary Figure S1). Moreover, the expressions of PPARγ, Fabp4, adiponectin, and perilipin A, which are adipogenic markers, were decreased in the CM-P10 group (Figures 2F,G). The adipogenic capacity was decreased in the CM-P10 group. Taken together, these data indicate that the proliferation and differentiation capacity of BMSCs is significantly decreased by treatment with a conditioned medium of senescent cells.

With aging, the external microenvironment has a greater impact on the proliferation capacity of the BMSCs. However, the TUNEL assay showed that the external environment in aging had no obvious effect on BMSC apoptosis (Supplementary Figure S2). This result shows that the external environment in aging mainly reduces the number of BMSCs by reducing the proliferation of the BMSCs rather than inducing apoptosis.

We cultured the BMSCs with CM-P3 and CM-P10 in vitro. The activity of β-galactosidase in the CM-P10 group was enhanced (Figure 3A). RT–PCR showed that the transcription level of the aging-related indicators P53, P21, and P16 was increased (Figure 3B). We know that the external microenvironment in aging can accelerate aging in young BMSCs. In addition, we measured the expression levels of Rex1, Nanog, and Pou5f1, which are related to cell stemness, and found that the external microenvironment in aging could reduce the stemness of BMSCs (Figure 3C). Western blot analysis verified that the protein expression level of SOX2, a dryness-related indicator, decreased after cells were cultured with CM-P10 (Figure 3D). These results demonstrated that the external environment in aging decreased the proliferation and stemness of the BMSCs and promoted BMSC senescence.

Chronic inflammation is associated with many diseases, and inflammatory factors play a key role. Studies have shown that elderly individuals are in a chronic, low-level subclinical pro-inflammatory state (Chung et al., 2011; Cevenini et al., 2013; Lin et al., 2017). To explore the relationship between the age-related inflammatory microenvironment and osteoporosis, we performed a Quantibody^® mouse inflammation array on serum extracts from young (aged 8 weeks) and old (aged 78 weeks) mice. The expressions of pro-inflammatory cytokines and chemokines were increased in the serum of old mice, and 14 cytokines reached statistical significance (Figure 4A). The cytokines with significant differences are shown in Supplementary Table S1. Among them, the pro-inflammatory factor IL-6 was increased, and the anti-inflammatory factor IL-10 was decreased most obviously, suggesting a trend of increased inflammation.

To evaluate the changes in IL-6 and IL-10 in the bone marrow microenvironment during aging, we studied femur sections from young (8-week-old) and aged (78-week-old) male C57BL/6 mice. We analyzed the levels of IL-6 and IL-10 in the bone marrow in different age groups by immunohistochemistry. The results showed an increase in IL-6 and a decrease in IL-10 in the aged group (Figures 4B,C). We confirmed this trend in CM by ELISA, and the results revealed increases in the pro-inflammatory cytokine IL-6 and decreases in the anti-inflammatory cytokine IL-10 in CM-P10 (,Figures 4D,E). In summary, we demonstrated that the age-associated shift in the inflammatory microenvironment is associated with an increase in the pro-inflammatory cytokine IL-6 and a decrease in the anti-inflammatory cytokine IL-10.

The age-related inflammatory microenvironment is characterized by an imbalance in inflammatory and anti-inflammatory pathways and is the basic mechanism of aging (Fougère et al., 2017; Franceschi et al., 2007). The GO and KEGG analyses (Figures 5A,B) were used to predict the potential mechanism. We hypothesized that these differentially expressed signaling pathways were the potential mechanism by which the BMSC functional changes were caused by the inflammatory microenvironment.