Mice

The plasmid pOG2-Cre (a gift from Xiao Yang, National Center for Protein Sciences, Beijing, China), which contains a 0.9 kb Bglap promoter to drive osteoblast-specific gene expression, was used to generate Bglap-HOTAIR transgenic mice (Bglap-HOTAIR TG). The plasmid pBS-mPrx1 (#13961, Addgene), which contains the 2.4 kb Prx1 promoter to drive limb mesenchymal stem cell-specific gene expression, was used to generate Prx1-HOTAIR transgenic mice (Prx1-HOTAIR TG). The Bglap-HOTAIR plasmid was constructed by inserting HOTAIR into the corresponding site downstream of the Bglap promoter. The Prx1-HOTAIR plasmid was constructed by inserting HOTAIR into the corresponding site downstream of the Prx1 promoter. These plasmids were linearized with KpnI, and then, the fragments of Bglap-HOTAIR/Prx1-HOTAIR were purified and microinjected into C57BL/6 J mouse embryos. The embryos were then surgically transferred into pseudopregnant B6D2F1 (C57BL/6 X DBA2) female mice as foster mothers at the Laboratory Animal Research Center of Tsinghua University. Animals were bred and maintained under specific pathogen-free (SPF) conditions in the Animal Research Building of the China Astronaut Research and Training Center (12 h light, 12 h dark cycles, temperature controlled at 21 ± 2 °C with free access to food and water). All the experimental procedures were approved by the Committees of Animal Ethics and Experimental Safety of the China Astronaut Research and Training Center (Reference number: ACC-IACUC-2020–006).

Immunohistochemistry and bone histomorphometry

The tibiae from mice were fixed with 4% PFA for 48 h, decalcified in 10% EDTA solution for 2 weeks, and then embedded in paraffin. The bone tissue was cut into approximately 5–7 μm thin slices by a microtome and absorbed on the slide. According to the IHC protocol, paraffin-embedded sections were dewaxed in xylene and rehydrated in gradient ethanol. Protease K was used for antigen retrieval at room temperature or 37 °C for 30 min, and then, a solution of 3.0% H2O2 was used to block the activity of endogenous peroxidase. The sections were blocked with goat serum and then incubated with an antibody against osteocalcin (1:500, Proteintech, 23418-1-AP) overnight at 4 °C. After three washes with PBS, horseradish peroxidase-labeled secondary antibodies were added and incubated at room temperature for 1 h, and then color development was performed with a DAB kit (ZSGB-bio). The sections were examined using a microscope (ECLIPSE CiS, Nikon).

Micro-CT

The distal femurs, cortical bone, and fifth lumbar vertebra bone of mice were scanned by a micro-CT system (SCANCO Medical μ40, Bruttisellen, Switzerland). For distal femurs, the region of trabecular bone proximal to the distal growth plate was selected for analyses within a conforming volume of interest commencing at a distance of 210 μm from the growth plate and extending a further longitudinal distance of 1 050 μm in the proximal direction. Cortical bone measurements were performed in the diaphyseal region of the femur starting at a distance of approximately 3.50 mm from the growth plate and extending a further longitudinal distance of 840 μm in the proximal direction. The trabecular region of the fifth lumbar vertebra bone was defined manually to exclude the cortical component. All trabecular bone from each selected slice was segmented for 3-D reconstruction to calculate the following parameters: bone volume per total volume (BV/TV), bone mineral density (BMD), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular space (Tb.Sp) and structure model index (SMI).

OVX mouse model

All the female mice used were maintained under standard animal housing conditions. The 3-month-old female mice were divided into ovariectomy (OVX) or sham-operated groups. At 3 months after OVX surgery, bone tissues and bone marrow cells were collected from sham-operated and OVX mice. All of the experimental procedures were approved by the Committees of Animal Ethics and Experimental Safety of the China Astronaut Research and Training Center.

Aged mouse model

Twenty-four-month-old male mice were purchased from Yangzhou Youdu Biotechnology Co., Ltd., and 2-month-old male mice were used as controls for young mice. Bone tissues and bone marrow cells were collected from young and aged mice. All of the experimental procedures were approved by the Committees of Animal Ethics and Experimental Safety of the China Astronaut Research and Training Center.

Three-point bending analysis

The femurs were stored in phosphate buffer immediately after being removed from the mice. The three-point bending test (span length, 4.0 mm; loading speed 0.50 mm·s−1) at the mid-femur was performed using Texture Analyzer Texture Pro CT V1.6 Build, Brookfield Engineering Labs, Inc.

Hindlimb unloading (HU) mouse model

Hindlimb unloading was performed to remove body weight-induced mechanical loading on hindlimbs, which mimicked the bone loss induced by weightlessness.62 The mice were individually caged and suspended by the tail using a strip of adhesive surgical tape attached to a chain hanging from a pulley. The mice were suspended at a 30° angle to the floor with only the forelimbs touching the floor; this allowed the mice to move and access food and water freely. The 3-month-old male mice were subjected to hindlimb unloading with tail suspension for 28 d. After euthanasia, bilateral femurs and tibiae were dissected and processed for cell sorting with a flow cytometer, micro-CT examination, bone histomorphometry analysis, and qPCR analysis.

Fluorescence activated cell sorting (FACS)

Whole bone marrow cells were flushed with α-MEM containing 10% penicillin‒streptomycin from the bone tissue of hindlimb unloading (HU, 3-month-old male mice were subjected to hindlimb unloading with tail suspension for 28 d), ovariectomy (OVX, 3-month-old female mice were OVX or sham-operated for 3 months) and aging model (24-month-old male mice) mice. Flow cytometry was used to isolate Sca-1+CD29+CD45−CD11b− BMSCs and ALP+ osteoblasts. Cells were sorted in TRIzol reagent for RNA extraction and qPCR analysis, and specific primers are listed in Table S2.

Measurement of serum PINP concentrations

The serum levels of PINP were detected by a mouse PINP (procollagen type I N-terminal propeptide) ELISA kit (ImmunoWay, KE1744) according to the product instructions.

Alkaline phosphatase staining and Alizarin red staining

For alkaline phosphatase staining, the cells were fixed with 4% paraformaldehyde (PFA) for 10 min and rinsed with PBS three times at room temperature. Alkaline phosphatase staining was monitored using a Vector Blue Substrate Kit (SK-5300; Vector Laboratories). According to the protocol, the cells were incubated with the substrate working solution for 20–30 min and then rinsed with PBS. For Alizarin red staining, cells were cultured in an osteogenic induction medium for 14 days, fixed with 4% PFA and stained with 40 mmol·L−1 Alizarin red S (ARS, Sigma-Aldrich, A-5533) at a pH of 4.0 for 15 min with gentle agitation. Then, the cells were rinsed five times with double-distilled H2O while gently agitating. Both staining procedures were protected from light.

Double calcein labeling

Peritoneal injection with calcein (30 mg·kg−1 body weight) was performed at 12 and 2 days before mouse euthanasia. The tibias were harvested for undecalcified histology analysis. Unstained 15 mm sections were examined using fluorescence microscopy. Statistical analyses were performed with the BioquantOsteo Analysis System.

Whole-mount staining

Skeletal preparations were stained with ARS and Alcian blue as previously described.63 In brief, newborns were eviscerated, and the skin was removed. The samples were fixed with 95% ethanol for 3 days and incubated with acetone for another 48 h. Subsequently, the samples were stained in Alcian blue solution for 3 days. Then, they were cleared with 75% ethanol three times for 1.5 h each, followed by treatment with 1.0% KOH overnight. After staining with 0.005% ARS solution for 5 h, the samples were cleared and conserved in 1.0% KOH/20% glycerol.

Cell culture and osteogenic differentiation

BMSCs were isolated from the bone marrow of 1-month-old mouse femurs and tibias. Briefly, the femurs and tibias were dissected from soft tissue and placed in PBS containing 10% penicillin‒streptomycin. After removal of the remaining muscle and connective tissue, the epiphyses were cut away. The bones were flushed with α-MEM containing 10% penicillin‒streptomycin to remove bone marrow cells. Cells were seeded in 10 cm plates for 2 days with α-MEM containing 10% fetal bovine serum (FBS; Gibco, Grand Island, USA), the suspended cells were removed with the replacement solution, and the adherent cells were used as BMSCs. For primary osteoblasts, cells were isolated from the calvaria of neonatal mice within 3 days. Briefly, mouse calvaria was dissected aseptically and digested with 0.2% collagenase, and then, the digested cells were seeded in culture flasks with α-MEM containing 10% FBS (Gibco, Grand Island, NY, USA) and 1.0% penicillin‒streptomycin (Gibco). For osteogenic differentiation, cells were cultured in an osteogenic induction medium with α-MEM containing 10% fetal bovine serum, an additional 50 μg·mL−1 ascorbic acid (Sigma, A4403), and 5 mmol·L−1 β-glycerophosphate (Sigma, G9422).

RNA extraction and qPCR analysis

Total RNA was extracted from osteoblasts or bone tissues with TRIzol reagent (Life Technologies, 15596018) according to the manufacturer’s instructions. RNA was reverse transcribed with the PrimeScript RT reagent Kit (TaKaRa, China) according to the manufacturer’s instructions. Quantitative reverse transcriptase-PCR (qPCR) was performed using a SYBR Premix Ex Taq II Kit (TaKaRa, China). The level of endogenous mRNA was normalized to the level of Gapdh mRNA using the 2−ΔΔCT method. Specific primers are listed in Table S2.

RNA fluorescence in situ hybridization and immunofluorescence microscopy

For specific probe synthesis, primers containing the T7 promoter were added to the HOTAIR template, and the amplification reaction was performed. The PCR products were recovered through agar gels and purified, and then, the purified PCR products were used as templates for in vitro transcription using a digoxigenin (DIG) labeling kit (Roche). RNA FISH was carried out as described in a previous study.64 Briefly, osteoblasts and BMSCs were rinsed gently with PBS, fixed in 4% fixative solution for 15 min at room temperature, and then permeabilized in 0.5% Triton X-100/PBS on ice for 10 min to increase the permeability of the cell membrane. Subsequently, the cells were incubated with prehybridization solution at 55 °C for 1.0 h and then incubated with a hybridization solution containing DIG-labeled HOTAIR or other RNA probes at 5–10 μg·mL−1 overnight at 55 °C. After hybridization, the cells were washed 3 times with prewarmed wash buffer at 55 °C for 10 min. After that, the cells were washed 3 times with buffer 1 (2× SSC and 0.01% Tween-20) at 55 °C for 10 min and then washed 3 times with buffer 2 (0.2× SSC and 0.01% Tween-20) at 55 °C for 10 min. Next, the cells were washed with 1× TBST for 5 min and incubated with 1× blocking buffer for 1 h at room temperature. Then, the cells were incubated with mouse anti-DIG antibody (1:200, Abcam, Cat. No. ab420, monoclonal) or rabbit anti-HuR antibody (1:300, Cell Signaling Technology, Cat. No. 12582) in 1× blocking buffer at room temperature for 1 h, washed 3 times with TBST for 5 min, incubated with secondary antibodies (1:300, FITC-labeled goat anti-rabbit IgG, ORIGEN, ZF-0311; 1:300, TRIC-labeled goat anti-mouse IgG, ORIGEN, ZF-0313) in 1× blocking buffer at room temperature for 1 h, and washed 3 times with 1× TBST for 5 min. Finally, cell nuclei were stained with Hoechst (Sigma-Aldrich) and washed with PBS 3 times. Cy3-labeled probes specific for 18 S RNA were used as cytoplasmic markers, and Cy3-labeled probes specific for U6 RNA were used as nuclear markers. The probe sequences for U6 and 18 S were as follows: 18 S: 5′-CTTCCTTGGATGTGGTAGCCGTTCC-3′, U6: 5′- GCAGGGGCCATGCTAATCTTCTCTGTATCG-3′. Images were taken with a Zeiss LSM 710 microscope for confocal scanning.

Structured illumination microscopy (SIM)

SIM images of BMSCs and osteoblasts were acquired on the DeltaVision OMX V3 imaging system (Cytiva, GE Healthcare) with a ×100/1.40 NA oil objective (Olympus UPlanSApo), solid-state multimode lasers (488 nm, 405 nm, 561 nm) and electron-multiplying CCD (charge-coupled device) cameras (Evolve 512×512, Photometrics). For optimal images, immersion oils with refractive indices of 1.512 were used for BMSCs and osteoblasts on glass coverslips. The microscope is routinely calibrated with 100 nm fluorescent spheres to calculate both the lateral and axial limits of image resolution. SIM image stacks were reconstructed using WoRx 6.1.1 software (Cytiva, GE Healthcare). Conventional image stacks were processed by deconvolution methods using WoRx 6.1 software (Cytiva, GE Healthcare). The reconstructed images were further processed for maximum-intensity projections with WoRx 6.1.1 software.

RNA immunoprecipitation

BMSCs and osteoblasts were seeded in 15 cm culture plates, harvested after reaching over 90% confluency, washed twice with ice-cold PBS and suspended in 1 mL of RNA immunoprecipitation (RIP) buffer (50 mmol·L−1 Tris pH 7.4, 150 mmol·L−1 NaCl, 0.5% Igepal, 2 mmol·L−1 VRC, 1 mmol·L−1 PMSF and 1× protease inhibitor cocktail) with RNase inhibitor (TaKaRa, 2313 A). After 30 min of incubation at 4 °C on a rotating wheel, the lysates were centrifuged at 1 000 × g for 10 min at 4 °C, and the supernatants were precleared with 20 μL of Protein G PLUS-agarose beads (Santa Cruz, sc-2002). The precleared supernatants were then equally divided into two parts and incubated with 30 μL of Protein G PLUS-agarose beads with rabbit anti-HuR antibody (1:300, Cell Signaling Technology, Cat. No. 12582, monoclonal) or rabbit IgG (1:500, Cell Signaling Technology, Cat. No. 3900, monoclonal) for 3 h at 4 °C, followed by washing three times with RIP buffer. Finally, TRIzol reagent was added to the precipitate, and RNA was isolated for real-time RT‒PCR. Primers used for template amplification are shown in Table S2.

RNA pulldown

Biotinylated full-length HOTAIR was synthesized using a T7 promoter-bearing PCR product that contained full-length HOTAIR, and biotinylated miR-214 was synthesized by Ruibo Biotechnology Co., Ltd. Cell lysates from the indicated BMSCs or osteoblasts were incubated with biotin-labeled full-length HOTAIR or miR-214 (1 μmol·L−1) for 2 h and then with 20 μL of streptavidin-coupled agarose beads for 1 h. After extensive washing, the precipitated complexes were analyzed by real-time RT‒PCR, Western blotting and protein spectrum. The primers used for template amplification are shown in Table S2.

MS2-based RNA immunoprecipitation (RIP) assay

The MS2-based RNA immunoprecipitation (RIP) assay was performed as described in previous studies.65 Osteoblasts were cotransfected with pcDNA3.1-12×MS2-Ctrl, pcDNA3.1-12×MS2-HOTAIR, and pcDNA3.1-MS2-YFP (Addgene). After 48 h, the transfected cells were collected for lysis using RIPA buffer (50 mmol·L−1 Tris-HCl [pH 7.4], 150 mmol·L−1 NaCl, 0.1% SDS, 1% NP-40, 1 mmol·L−1 PMSF, 1× proteinase inhibitor cocktail and 1× phosphatase inhibitor [P1260]). The lysate was incubated with rabbit anti-GFP antibody (1:200, Abcam, Cat. No. ab290, polyclonal) or rabbit IgG (1:200, Abcam, Cat. No. ab172730, monoclonal) overnight. The RNA/protein complex was recovered with protein G Dynabeads and washed with RIPA buffer several times. RNA was recovered with TRIzol and analyzed by RT‒PCR.

Cell transfection

Plasmid and siRNA transfection was carried out with Lipofectamine 3000 (Invitrogen, L3000-015) or RNAi Max (Invitrogen, 13778-150) according to the manufacturers’ protocol. The siRNA sequences used in the study were as follows: mouse negative control (NC) siRNA: 5′-AACGUACGCGGAAUACUUCGA-3′; mouse HOTAIR: 5′-GCAGAAUUCACUCUCAAUATT-3′; and mouse HuR: 5′-AAGAGGCAAUUACCAGU UUCA-3′.

Western blotting

BMSCs or primary osteoblasts were lysed in buffer (150 mmol·L−1 NaCl, 50 mmol·L−1 Tris, pH 7.8, 10% glycerol, 1 mmol·L−1 EGTA, 0.5% NP-40, 1 mmol·L−1 EDTA, 1 mmol·L−1 PMSF, 1x cocktail) on ice with vigorous shaking for 30 min. The lysates were subjected to Western blotting with the indicated antibody. The following antibodies were used in the study: rabbit anti-HuR antibody (1:1 000, Cell Signaling Technology, Cat. No. 12582, monoclonal), rabbit anti-Atf4 antibody (1:1 000, Cell Signaling Technology, Cat. No. 11815, monoclonal), and rabbit anti-Gapdh antibody (1:5 000, Abways, Cat. No. AB0036, monoclonal).

ChIP assay

ChIP assays were performed using a commercial kit from Cell Signaling Technology (Cell Signaling Technology, Cat. No. 56383). Briefly, BMSCs transfected with si-HOTAIR or NC were fixed at room temperature with 1% formaldehyde for 10 min and quenched for 5 min with glycine solution. The nuclei of BMSCs were extracted, and chromatin was fragmented into lengths of 0.2-1.0 kb by ultrasound. The chromatin fragments were then immunoprecipitated with the ChIP-validated rabbit anti-H3K27me3 antibody (1:200, Cell Signaling Technology, Cat. No. 9733, monoclonal), rabbit anti-H3 antibody (1:200, Cell Signaling Technology, Cat. No. 4499, monoclonal), or rabbit IgG (1:500, Cell Signaling Technology, Cat. No. 3900, monoclonal) overnight at 4 °C, followed by incubation at 4 °C with ChIP-Grade Protein G Magnetic Beads for 2 h. After washing, the eluted DNA was purified and assessed by qRT‒PCR to measure the amount of enrichment of the region of the Runx2 or Sp7 gene promoter. Histone H3 antibody pulldown for enrichment served as a positive control. The primers used for qRT‒PCR are shown in Table S2.

Statistical analysis

For statistical analysis, all quantitative data are presented as the mean ± SEM. Student’s t test was used for statistical evaluations of two group comparisons. Statistical differences among groups were analyzed by 1-way analysis of variance (ANOVA) or 2-way ANOVA (if there were 2 factor levels), followed by a post hoc test to determine group differences in the study parameters. All statistical analyses were performed with Prism software (GraphPad Prism for Windows, version 8.0, Nashville, USA). Differences were considered significant at *P < 0.05, **P < 0.01, and ***P < 0.001.