Experimental Animals
All animal experimentation conducted in this investigation received approval from the Animal Ethics Review Committee of Shenzhen People's Hospital, and all procedural interventions strictly adhered to its stipulated guidelines. The investigators responsible for histological analyses were blinded to group information. Mice were randomly assigned to the experimental groups.
Generation of conditional Foxk1 knock out Mice
Foxk1 fl/+ mice were obtained from Professor Yu Nie of Fuwai Hospital, Chinese Academy of Medical Sciences. Osterix-Cre mice were purchased from BIOCYTOGEN (Shanghai, China). Ocn-Cre mice purchased from Cyagen Biosciences (Suzhou, China). Osterix-Cre and Ocn-Cre mice were crossed with Foxk1fl/+ mice for more than three generations to obtain Osx-cre;Foxk1fl/fl and Ocn-cre;Foxk1fl/fl conditional knockout mice. Tail genomic DNA is applied to identify the genotypes by PCR amplification. Primers for the Foxk1 knockout allele and Cre transgene genotyping are listed in Table S1. All mice were propagated and raised in specific pathogen free facilities.
MicroCT
The right femur of sacrificed mice was fixed in 4% paraformaldehyde and then scanned with a MicroCT system (SkyScan 1276, Bruker, Belgium) with a voltage of 80 kVp, current of 100 µA, exposure time of 926 ms, and a pixel size of 20 µm. Two-dimensional image of distal femur and three-dimensional image of trabecular bone of distal femur were constructed using NRecon software. The volume of interest (VOI) of femurs was defined as a cylindrical area at the distal region of the growth plate. The bone volume (BV), bone volume/tissue volume ratio (BV/TV), bone surface area/tissue volume (BS/TV), trabecular number (Tb.N) and trabecular separation (Tb.Sp) were analyzed.
ELISA
Blood collected from mice was centrifuged at 2,000g for 30 min to harvest the plasma. The concentration of C-terminal telopeptides of type I collagen (CTX-1) (Jianglai Biology, JL20123, Shanghai, China) in the plasma was analyzed according to the manufacturer’s instructions.
Bone biomechanical properties
The biomechanical properties were tested by a three-point bending mechanical strength device (AG-IS, Shimadzu, Japan) as previous described34. Briefly, the left femur was obtained from mice and cleaned of adherent tissue. Three-point bending test was performed at the mid-shaft of the femur wrapped by saline-soaked gauze. The femur was placed at two points 6mm apart, and the test specimen was moved down at a speed of 1mm/min. The mechanical data were generated and analyzed to determine the max force, max bending strength, and elastic modulus.
Bone histomorphometry
Mice were double labeling with calcein (20 mg/kg; Sigma-Aldrich, C0875, Missouri, USA) through intraperitoneal injection at 9 days and 3days before sacrificed for dynamic analysis. The right femurs were dehydrated with 20% sucrose solution and performed frozen sectioning at a thickness of 10 µm to obtain frontal sections of the trabecular bone using a Leica SM2500E microtome (Leica Microsystems, Wetzlar Germany). Obtain fluorescence images using an inverted fluorescence microscope (Leica image analysis system, Q500MC). Then the dynamic bone formation parameters, including mineral apposition rate (MAR) and bone formation rate (BFR), were analyzed by ImageJ software (NIH, USA).
Immunohistochemical analyses
The femur was fixed in 4% paraformaldehyde for 48 hours and then decalcified in 10% EDTA for 2 weeks. Decalcified bones are dehydrated and then embedded in paraffin. Bone specimens were cut into 5-µm-thick sections with a paraffin microtome (Leica, Wetzlar, Germany). The decalcified bone specimens were applied to perform H&E staining, IHC staining, IF staining and Trap staining.
IHC staining and IF staining were performed according to the protocol. Briefly, slides were dewaxed and then incubated the primary antibody overnight at 4℃. After incubating the secondary antibody at room temperature for one hour, take photos for testing. IHC staining for OCN was used to assess the number of OCN positive osteoblasts as previous described35. The numbers of osteoblasts on cancellous bone surface in the entire region of interest were counted and the data were then normalized to the number of osteoblasts per millimeter of trabecular bone surface perimeter (N mm− 1). Trap staining was used to assess the number of osteoclasts as previous described35. IHC and Trap staining images were obtained by light microscopy (TS2-S-SM; Nikon, Tokyo, Japan). IF staining images were obtained by an inverted fluorescence microscope (Leica image analysis system, Q500MC). FOXK1 (ab18196, 1:500) and GFP (ab290, 1:200) primary antibody were purchased from abcam (Cambridge, England). OCN (sc-390877, 1:200) and OSTERIX (sc-393325, 1:200) primary antibody were purchased from santa cruz biotechnology (California, USA). Anti-rabbit (A-11008, 1:400; A-11012, 1:400) and anti-mouse (A-11005, 1:400; A-11001, 1:400) secondary antibody was purchased from Invitrogen (California, USA). TRAP staining kit (387A) was purchased from Sigma Aldrich (Missouri, USA).
AAV9 construction and injection in mice
Adeno-associated virus-9 (AAV9) was used as a vector to overexpressed Foxk1 in mice. AAV9 vector was modified by preosteoblast specific promoter osterix to target exclusively osteoblast lineage cells, which was designed and synthetized by WZ Biosciences Inc. (Shangdong, China). 4×1011 GC AAV9 was diluted with PBS to 100ul and injected into each mouse through the tail vein. Tail vein injection of 100ul PBS as control group to perform Whole body and individual organ imaging of GFP expression and obtain tissues 2 weeks after injection. To investigate the role of glycolysis in Foxk1 mediated bone formation, 2DG (200uL, 0.5mg/g, Sigma-Aldrich, D8375) or PBS (200ul) was injected every three days via intraperitoneal until the mice are euthanized at 8 weeks after virus injection according previous report36.
Human samples
The bone samples of 12 man subjected to spinal fusion surgery in Shenzhen People’s Hospital were employed in this study. The inclusion criteria are based on bone density manifested by T-score. Individuals with low bone mineral density (T score ≤ -2.5 SD) were included in the osteoporosis group (n = 6), while those with normal bone mineral density (T score ≥ -1.0 SD) formed the control group (n = 6)39. Individuals with other bone diseases, such as rheumatoid arthritis and osteoarthritis, were excluded from the study. The study was approved by the research ethics committee of Shenzhen People’s Hospital. Each participant has signed an informed consent form.
Cell culture and osteoblast differentiation
Primary calvarial osteoblasts were isolated from mouse calvaria of 1 to 3-day-old pups. Briefly, the collected calvarias were cut into pieces and incubated with 2.5% tryptase (Gibco, 25200072, New York, USA) at 37°C with gentle shaking for 10 min. Supernatant was discarded and type I collagenase (Biyotime, ST2294-100mg, Shanghai, China) was added to cover the calvaria at 37°C with gentle shaking for 30 min. Supernatant was discarded again and type I collagenase was added to cover the calvaria at 37°C with gentle shaking for 60 min. Then supernatant was collected and centrifuged 300g for 5 minutes at room temperature. cell precipitation was resuspended using α-MEM (Gibco, 12571063) with 10% fetal bovine serum (Gibco, 10099141C) and maintained in an incubator under 5% CO2 at 37°C overnight. The next day, replace the culture medium with a new one to remove non adherent cells.
For all experiments, the primary calvarial osteoblasts were incubated with osteogenic induction medium, as previous described26, containing 10 mM b-glycerol phosphate (Sigma-Aldrich, G9422) and 50 ug/ml ascorbic acid (Sigma-Aldrich, A4544) and replace the culture medium every other day.
siRNA transfection and virus infection
The small interfering RNAs against Foxk1 (Table S) and control siRNAs were designed and synthesized by Hanbio Co. Ltd (Shanghai, China). siRNAs were transfected into the primary calvarial osteoblasts after reaching ~ 80–90% confluency using Lipofectamine 2000 (Invitrogen, 11668027, California, USA) in accordance with the manufacturer's instructions.
The adenovirus of overexpression of Foxk1 and control adenovirus in cells were designed and synthesized by WZ Biosciences Inc.. Adenovirus were added to the primary calvarial osteoblasts after reaching ~ 80–90% confluency with 2DG (2mM) or PBS.
western blotting
To isolate protein, cells were scraped, or bones were pulverized in RIPA lysis buffer (Biyotime, P0013B) containing protease inhibitors (Biyotime, P1005). Protein concentration was measured using the BCA method and 10 µg of the protein was run on 12–14% SDS-PAGE gels (Yeasen Biotechnology, 36249ES10, Shanghai, China) and transferred to PVDF membranes (Millipore, IPVH00005, Massachusetts, USA). Then the PVDF membranes were blocked with TBST containing 5% dried skimmed milk and incubated with anti-FOXK1 (ab18196, 1:1000), anti-RUNX2 (1:500, Proteintech Group, 20700-1-AP, Wuhan, China), anti-OSTERIX (sc-393325, 1:500), anti-ALP (1:500, Proteintech Group, 60294-1-Ig) and anti-β-Tubulin (1:1000, Proteintech Group, 10094-1-AP) at 4℃ overnight. Incubate the membranes containing the primary antibody with the secondary antibody at room temperature for 1 hour. Finally, the membranes were visualized using SuperSignal™ West Pico PLUS Chemiluminescent Substrate Reagent (Thermo Fisher Scientific, 34580 Massachusetts, USA).
qPCR
Total RNA was extracted from cells using the RNeasy kit (Qiagen, 74104, Hilden, Germany) and from organs using TRIzol reagent (Invitrogen, 15596018CN) according to the manufacturers’ instructions. The RNA concentration of all samples was examined using a Nano-Drop ND-2000 spectrophotometer (Thermo Fisher Scientific). Total 1000ng RNA was performed Reverse transcription using the Prime Script™ Reverse Transcript Master Mix (TaKaRa Bio, RR037A, Otsu, Japan). qPCR was performed on a StepOnePlus Real-Time PCR system (Applied Biosystems, Waltham, USA), and the relative mRNA concentrations normalized to that of β-actin were calculated by the 2−ΔΔCt method.
ALP and Alizarin red staining
ALP and Alizarin red staining was performed to examine the osteogenic ability of primary calvarial osteoblasts. Briefly, treated primary calvarial osteoblasts were induced osteogenic differentiation for 7days to perform ALP staining and for 14days to perform Alizarin red staining. Cells were washed twice at the corresponding time point and fix them with 4% paraformaldehyde for 10min. Then ALP staining solution (Beyotime, P0321S) or Alizarin red S staining solution (Beyotime, C0140-100ml,) at pH 8.3 was used to stain the cells for 30 min. After washing twice, the cells were photographed by light microscopy (TS2-S-SM, Nikon, Japan).
Glucose uptake assays and lactate measurements
Glucose uptake of primary calvarial osteoblasts was detected by a Glucose uptake kit (abcam, ab136955) according to the instructions. Briefly, primary calvarial osteoblasts were seeded to 96-well plates and starved for 4 hours with α-MEM without FBS. Then, Cells were incubated with 10mM 2DG for 20min. Cells were Lysed after wash twice and interacted with reaction mix. Measure output at OD 412 nm on a microplate reader (TECAN, Spark 10M, Switzerland). Uptake of glucose was normalized to the cell numbers.
Lactate levels of culture medium were measured using the Lactic Acid assay kit (Nanjing Jiancheng Bioengineering Institute, A019-1-1, Nanjing, China).
Extracellular acidification rate and oxygen consumption rate measurements
XF24 Extracellular Flux Analyzer (Seahorse Bioscience, Maine, USA) was used to measure the Extracellular Acidification Rate (ECAR) and Oxygen Consumption Rate (OCR) as previously described37. Briefly, 70000 primary calvarial osteoblasts were seeded in an XF24 V7 microplate (Seahorse Bioscience) coated with poly-L-lysine. Next, primary calvarial osteoblasts were treated with siNC, siFoxk1, or adenovirus for Foxk1 overexpression as described in the results. On the day of measurement, osteoblasts were washed twice with XF medium and incubated for 1 hour in a CO2 free incubator at 37°C before loading. Glucose (25mM), oligomycin (1µM), FCCP (2µM), antimycin A (0.5µM), rotenone (0.5µM), 2-deoxy-glucose (50mM) and BSA:palmitate (200µM) were used to measure ECAR (mpH/min) and OCR (pmol/min). Data were normalized to the cell numbers.
Metabolomics analysis
Primary calvarial osteoblasts were plated on 10cm cell culture dish and knocked down Foxk1 using siRNA. Cells were harvested to detect energy metabolism related metabolites using Acquity-I Xevo TQ-S LCMs (Waters, Massachusetts, USA). Use MassLynx software (v4.1, Waters) to process the raw data files generated by UPLC-MS/MS, and perform peak search, integration, calibration, and quantification for each metabolite.
Cell proliferation detection
Primary calvarial osteoblasts were seeded in 24 well plate size cell crawling plate. Next, primary calvarial osteoblasts were treated with siNC, siFoxk1, or adenovirus for Foxk1 overexpression as described in the results. Then cells were collected to perform Edu incorporation (Yeasen Biotechnology, 40275ES60, Shanghai, China) according to the instructions or IF staining of Ki67 (abcam, ab16667) to evaluate osteoblast proliferation. Obtain fluorescence images using an inverted fluorescence microscope (Leica image analysis system, Q500MC).
CUT&Tag
Primary calvarial osteoblasts were seeded in 12 well plate and overexpressed Foxk1 using adenovirus. Cells were collected to perform CUT&Tag using a Hyperactive Universal CUT&Tag Assay Kit for Illumina (Vazyme Biotechnolegy, TD903-01, Nanjing, China) after 72 hours treatment according to the manufacturer’s protocol. Briefly, Cells were collected and resuspended with 100ul wash buffer. Cell suspension incubated with activated ConA Beads for 10min at room temperature. The supernatant was discarded after instantaneous centrifugation and resuspended with 50µl precooled antibody buffer containing 1ug anti-FOXK1 antibody (abcam, ab18196). Cell suspension was placed at 4℃ overnight. The supernatant was discarded and sediment incubated with 50µl Dig-wash Buffer containing secondary antibody for 60min at room temperature with rotation. The supernatant was discarded and sediment incubated with 100µl Dig-300 Buffer containing 2µl pA/G-Tnp for 60min at room temperature. The supernatant was discarded and sediment incubated with 50µl Dig-300 Buffer containing 10µl 5×TTBL for 60min at 37℃ to make fragmentation of chromatin. Chromatin fragmentation incubated with 5µl Proteinase K, 100µl Buffer L/B and 20µl DNA Extract Beads for 10min at 55℃. The supernatant was discarded and beads were washed twice with Buffer WA and Buffer WB respectively. The supernatant was discarded and DNA eluted with 22µl sterilized ultrapure water. Then, Libraries were amplified and sequenced on Illumina Novaseq platform at Novogene Science and Technology Co., Ltd (Beijing, China) and 150 bp paired-end reads were generated.
Data obtained from CUT&Tag of Foxk1 was analyzed as described previously with modifications38. Raw reads of fastq format were firstly processed using fastp (version 0.20.0) to clean reads. Clean reads obtained from CUT&Tag of Foxk1 were aligned to the reference genome using BWA (v0.7.12) mem. Only uniquely mapped (MAPQ ≥ 13) and de-duplicated reads were used for further analysis. All peak calling was performed with MACS2 (version 2.1.0) and peaks with q-value threshold of 0.05 was used for all data sets. The position of peak summit around transcript start sites of genes can predict the interaction sites between protein and gene. ChIPseeker was used to retrieve the nearest genes around the peak and annotate genomic region of the peak. KOBAS software was used to test the statistical enrichment of peak related genes in KEGG pathways. Visualization of peak distribution along genomic regions of interested genes was performed with IGV.
ChIP assay
Primary calvarial osteoblasts were seeded in 15cm cell culture dish and overexpressed Foxk1 using adenovirus. Cells were collected to perform ChIP using a SimpleChIP® Enzymatic Chromatin IP Kit (Cell Signaling Technology, 9003, Massachusetts, USA) after 72 hours treatment according to the manufacturer’s protocol. In brief, 37% formaldehyde solution was used to crosslink protein and DNA of each sample. SDS lysis buffer with protease inhibitor cocktail was added after cell harvesting. Chromatin was digested by enzyme and ultrasound. Chromatin fragments were incubated with 1ug primary antibody of anti-FOXK1 (abcam, ab18196) or anti-IgG (Cell Signaling Technology, 9003) at 4℃ overnight. DNA samples were precipitated with beads and quantified with specific primers (Table S3) via qPCR.
Statistical analysis
Statistical analysis was performed using Prism 8 (GraphPad software) and data are presented as mean ± SEM. Statistical significance was determined with the unpaired two-tailed Student’s t-test. P < 0.05 was considered significant.