Close this search box.

Histone demethylase KDM7A regulates bone homeostasis through balancing osteoblast and osteoclast differentiation – Cell Death & Disease


The Osx-cre mouse which targets Osx-expressing osteoprogenitor cells was purchased from Biocytogen (Beijing, China;; #110131). It was generated by placing an F2A-iCre sequence between exon 2 and 3’UTR of osterix gene in C57BL/6 embryonic stem cells. The Kdm7a floxed mouse which contains LoxP sites flanking exon 2 of mouse Kdm7a was generated via CRISPR/Cas9 technology by GemPharmatech (Nanjing, China). Kdm7afx/fx mice were crossed with Osx-cre mice to generate the knock-out line Osx-cre; Kdm7afx/fx (Kdm7a cKO). The sequences of primers used for genotyping are listed in Supplemental Table 1.

To investigate the bone phenotypes of Kdm7a cKO mice, Kdm7afx/fx mice were used as controls. To investigate whether deletion of Kdm7a in osteoprogenitor cells prevents ovariectomy (OVX)-induced bone loss, 12-week-old female cKO and Kdm7afx/fx mice were subjected to OVX or sham operation under anesthesia. For each genotype, the mice for OVX and sham operations were randomly allocated. Twelve weeks after surgery, tibial samples were collected and bone morphometric parameters were analyzed via μCT followed by histological staining.

Animal experiments were carried out the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines [40]. The sample size for the animal experiments was determined based on a survey of data from published papers. Five mice were housed per cage in a standard individually ventilated cage (IVC) system (Tecniplast, Italy) in an SPF facility maintained under controlled temperature (22 °C) and humidity (50–60%), as well as a 12 h dark:12 h light cycle. The mice were given free access to autoclaved water and food. All the experimental animals were located in the same animal room and subjected to the same environmental conditions to avoid confounding factors. All animals with correct genotypes were included in data collection. The animal protocol and experimental methods were approved by the Animal Ethics Committee of Tianjin Medical University and Chu Hsien-I Memorial Hospital.

μCT analysis

Tibial and lumbar vertebral samples from 24-week-old mice were scanned using a Scanco viva CT80 (Scanco Medical AG, Switzerland) operated with the following parameters: voxel size of 10.4 μm, energy at 55 kVp, intensity of 145 μA, FOV/Diameter of 31.9 mm, and 300 msec integration time. The measurements were quantitatively analyzed. Three-dimensional reconstructions were made using Scanco software. The region of interest for histomorphometric parameters was selected in the 1 mm region starting 0.2 mm below the growth plate.

Cell cultures

Primary BMSCs were isolated from the marrow of 6-week-old mice and cultured following a previously described procedure [41]. Pre-osteoblastic cells were isolated from the calvaria of 3-day-old mice. Briefly, the calvaria were minced into 1 mm×1 mm tissue blocks and subjected to five rounds of digestion using an enzyme mixture consisting of 0.25% trypsin and 0.1% type I collagenase at 37 °C. The cells collected from the last 4 digestions were pooled, centrifuged, resuspended and cultured in α-MEM supplemented with 10% FBS.

The cells were passaged at 80% confluence and those at passages 3–5 were induced to allow osteogenic and adipogenic differentiation following a previously published protocol [41]. In certain experiments aimed at investigating the underlying mechanisms, cells were treated with 200 ng/ml recombinant FAP (Sino Biological), or TC-E 5002 (MedChemExpress, USA), a selective inhibitor of the KDM2/7 subfamily, at doses of 10, 20 and 40 μmol/l.

Oil red O staining

After 5–6 days of adipogenic differentiation, differentiated adipocytes were fixed with 4% paraformaldehyde for 10 min, washed with PBS, and rinsed with 60% isopropanol. The cells were stained with 0.24% fresh oil red O solution for 5 min. To quantify the intensity of the staining, the oil red O dye retained in the cells was extracted using isopropanol, and the optical density was measured at 520 nm.

Staining of differentiated osteoblasts

Differentiated osteoblasts were subjected to ALP staining after 14 days of osteogenic induction, or to alizarin red staining after 21 days of induction as per a previously published procedure [41]. To quantify the intensity of the alizarin red staining, the stained cultures were destained with a solution containing 0.5 mol/L HCl and 5% SDS for 10 min, and the absorbance of the solution was read at 415 nm.

ALP activity measurement

Cell extract was diluted 5-fold and incubated with para-nitrophenyl phosphate (pNPP) liquid substrate (Beyotime Biotech, Shanghai, China) for 5 min at 37 °C. The absorbance was measured at OD405 on a microplate reader after addition of the termination solution. For normalization, total protein concentration was measured using a BCA protein assay kit (EpiZyme, Shanghai, China) and OD562 was measured. The OD405 value was divided by OD562 value to generate a standardized value for each sample.


Total RNA was extracted from cultured cells using a kit (Omega Biotek, Norcross, GA) or from tissues using RNAiso (Vazyme, Nanjing, China). cDNA synthesis and quantitative PCR amplification were performed using PerfectStart Uni RT&qPCR Kit (TransGen, Beijing, China). The expression levels of target genes were calculated by comparative Ct (ΔΔCt) method using β-actin as an internal reference. The primer sequences are listed in Supplemental Table S1.


Total RNA was isolated from calvarial cells of Kdm7a cKO mice and Kdm7afx/fx mice. The RNA-seq transcriptome library was prepared with 1 μg of total RNA, and subsequently sequenced on the MGISEQ-2000 platform (sequencing length: PE150) by Beijing Genomics Institute (Wuhan, China). Gene Ontology functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were conducted.

Western blotting

Cells were lysed in RIPA lysis buffer and protein concentration was measured using a BCA assay kit (EpiZyme). The proteins were separated via 10% SDS-PAGE, and then transferred onto nitrocellulose membranes. The membranes were incubated with primary antibodies at 4 °C overnight, followed by incubation with horseradish peroxidase-conjugated secondary antibodies. Protein bands were detected with a chemiluminescence kit (Epizyme). The primary antibodies used are listed in Supplemental Table S2.

Chromatin immunoprecipitation (ChIP) assay

ChIP assay was performed using an ABclonal kit (Wuhan, China). Briefly, mouse BMSCs were incubated with 1% formaldehyde for cross-linking, and then sonicated to generate 200–500 bp genomic DNA fragments. The cell lysates containing chromatin complexes were incubated overnight at 4 °C with 5 μg of anti-KDM7A (Biorbyt, orb67004), anti-H3K9me2 (ABclonal, A2359), anti-H3K27me2 antibody (ABclonal, A2362) or IgG, and then incubated with protein A/G beads to capture the immunocomplexes for 2 h at 4 °C. The de-crosslinked and purified DNA was used as a template to PCR-amplify mouse Rankl or fibroblast activation protein α (Fap) promoter sequences. Data are expressed as the percentage of input DNA. The primer sequences for ChIP-qPCR are listed in Supplemental Table S1.

Osteoclast differentiation from bone marrow cells

Bone marrow cells isolated from the long bones of 6-week-old Kdm7a cKO mice or Kdm7afx/fx mice were seeded into a 48-well plate at the density of 4×105 cells/well. The cells were cultured in α-MEM supplemented with 10% FBS and 20 ng/ml M-CSF (Sino Biological, Beijing, China) for 3 days, and then induced with 20 ng/ml M-CSF and 5 × 10−7 mol/l 1,25(OH)2D3 (APExBIO, USA) to allow osteoclast differentiation. The medium was refreshed every 3 days. After 8 days of induction, the cultures were stained with a TRAP staining kit (Sigma-Aldrich, USA). TRAP-positive multinucleated osteoclasts (≥3 nuclei) were counted.

Osteoclast differentiation from co-cultures

Bone marrow cells were collected from 10- to 14-day-old C57BL/6 J mice and cultured in α-MEM containing 10% FBS. After 24 h, the nonadherent BMM cells were collected.

For cell-cell contact co-culturing, BMM cells were seeded into a 24-well plate at the density of 5×105 cells/well. After 3 days of culturing in α-MEM supplemented with 10% FBS and 20 ng/ml M-CSF, the BMSCs from Kdm7a cKO or Kdm7afx/fx mice were seeded into the cultures at the density of 5,000 cells/well. The co-cultures were induced with 20 ng/ml M-CSF and 5 × 10−7 mol/l 1,25(OH)2D3 to allow osteoclast differentiation.

Transwell co-culturing was conducted using 24-well transwell plates with 0.4 μm membrane pores (Biofil, Guangzhou, China). Briefly, BMM cells were seeded into the transwell bottom compartment at the density of 5×105 cells/well and cultured in the presence of 20 ng/ml M-CSF. The BMSCs from Kdm7a cKO or Kdm7afx/fx mice were seeded into the top compartment at the density of 5,000 cells/well and cultured in the presence of 5 × 10−7 mol/l 1,25(OH)2D3. The medium was refreshed every 3 days. After 8 days of induction, TRAP staining was performed and osteoclasts were counted.

Enzyme-linked immunosorbent assay (ELISA)

ELISA was conducted following the manufacturer’s protocols to measure the levels of serum CTX-1 (SAB, College Park, MD, USA), RANKL (SAB), and PINP (SAB).

Histological staining

Bone samples were fixed with 10% formalin for 3 days, decalcified with 14% EDTA (pH 7.4) for 21 days, and embedded in paraffin. Then the samples were cut into 4-μm thick sections. Deparaffinized and rehydrated sections were stained with standard ABH/OG and hematoxylin/eosin (H&E). Marrow adipocyte number was counted and adipocyte area was measured. To identify osteoclasts, TRAP staining was performed using staining solution containing 0.1 mol/l sodium acetate (pH 5.0), 0.5 mol/l L-(+) tartaric acid, 0.2 mmol/l naphthol AS-MX and 1.5 mmol/l fast red violet LB salt. TRAP-positive osteoclasts were counted. The region of interest is 1 mm long starting from 0.2 mm below the growth plate.

Immunohistochemical staining

For antigen repair, the deparaffinized and rehydrated sections were digested with 0.075% trypsin at 37 °C for 15 min. After removing endogenous peroxidase with 3% H2O2 and blocking with 1% BSA, the sections were probed with primary antibodies overnight at 4 °C, and then incubated with HRP-conjugated secondary antibodies for 2 h at 4 °C. The primary antibodies used included anti-ALP (Huabio, Hangzhou, China), anti-RANKL (WL00285, Wanleibio, Shenyang, China) and anti-FAP (WL04890, Wanleibio). A 3,3’-diaminobenzidine (DAB) staining kit (GeneTech, Shanghai, China) was applied to visualize the staining. The nuclei were counterstained with hematoxylin.

Dynamic bone histomorphometry

Twenty-four-week-old female mice were injected intraperitoneally twice with calcein (Sigma, USA) at the dose of 10 mg/kg 2 and 7 days, respectively, prior to sacrifice. The femur samples were dehydrated in a graded series of ethanol and embedded, non-decalcified, in methyl methacrylate resin and sectioned at 8-μm thickness using a microtome (Leica RM2155, Wetzlar, Germany). The region of interest is 1 mm long starting from 0.2 mm below the growth plate. The unlabelled perimeter, single-labelled perimeter, double-labelled perimeter, and the area between the double labels were measured using the OsteoMeasure software (OsteoMetrics, Atlanta, GA, USA). Histomorphometric parameters, including MAR, MS/BS, and BFR/BS were calculated.

Data illustration and statistical analysis

Cell-based experiments were performed independently 5 times. The relative levels of mRNAs and proteins in the control group were set to 1. Statistical analyses were performed using GraphPad Prism 8.0 in a blinded manner. Data are presented as box-and-whiskers plots in which the boxes extend from the 25th to 75th percentiles, and the lines in the middle of the boxes indicate medians; all individual values are shown and the whiskers represent the minimum and maximum values.

Data were initially subjected to normal distribution analysis using Kolmogorov-Smirnov nonparametric test. For the comparisons between two groups, two-tailed unpaired independent Student’s t test was conducted, with Welch’s correction when there was unequal variance. For comparisons among multiple groups, One-way ANOVA followed by Dunnett’s test or two-way ANOVA followed by Tukey’s test was conducted. P < 0.05 was regarded as significantly different.