DZNep Promotes Mouse Bone Defect Healing Via Enhancing Both .

Cao et al. Stem Cell Research & Therapy(2021) 12:605RANKL and M-CSF were obtained from R&D (Minneapolis, MN, USA). The cell counting kit-8 (CCK-8) waspurchased from Dojindo Molecular Technology (Japan).The TRAP staining kits were bought from Sigma-Aldrich(St. Louis, MO, USA). The SYBR Premix Ex Taq IIand Prime Script RT reagent Kit were obtained fromTakara Biotechnology (Otsu, Shiga, Japan). The primaryantibodies against β-actin, LaminB, ERK, phospho-ERK(Tr202/Tyr204), JNK, phospho-JNK (Tr183/Tyr185),p38, phospho-p38, IKKβ, IκBα, P65, phospho-IKKα/β,phospho-IκBα, phospho-P65, EZH2, H3K27me3,GSK-3β, phospho-GSK-3β(ser9), and the secondary antibody, were bought from Cell Signaling Technology (CST,Danvers, MA, USA). The primary antibodies againstNFATc1, c-FOS were purchased from Absin BioscienceInc (Shanghai, China). The primary antibodies againstFoxc1, Wnt4 were purchased from Abcam (Cambridge,UK). Except β-actin is mouse anti-mouse. All the antibody mentioned are rabbit anti-mouse.Bone marrow‑derived macrophages (BMMs) preparationand cytotoxicity assayPrimary BMMs were obtained from whole bone marrow of male 6-week-old C57BL/6 mice. In brief, cellswere extracted from the tibiae and femurs bone marrowand suspended in complete α-MEM (α-MEM supplemented with 30 ng/ml M-CSF, 10% FBS, and 1% penicillin/streptomycin). Then, the cell cultures were kept ina humid environment at 37 C with 5% CO2 until theyreached 90% confluence. The cytotoxic effects of DZNepon BMMs were then determined by the CCK-8 kit. Specifically, the BMMs were seeded into 96-well plates intriplicate at the density of 8 103 cells/well in presenceof 100μL complete α-MEM for 24 h. After that, the cellswere treated with increasing concentration of DZNep(control, 3.125, 6.25, 12.5, 25, 50, 100, 200 nM) for 24, 48,and 96 h. After treatment, 100μL medium contained 10%CCK-8 buffer was added to the wells and incubated in thedark at 37 C for 2 h. The absorbance was then detectedat 450 nm wavelength (650 nm reference) on a microplatereader.Osteoclast differentiation and TRAP staining assayTo evaluate the role of DZNep on osteoclastogenesis,BMMs were seeded into, in triplicate, a 96-well plate atthe density of 8 103 cells/well with complete α-MEMfor one day. Afterward, the culture medium was replacedby the complete α-MEM containing different concentrations of DZNep (control, 3.125, 6.25, 12.5, and 25 nM)and RANKL (50 ng/ml) to stimulate osteoclast differentiation. Cells untreated with DZNep were includedas the control group. Moreover, the BMMs treated withor without DZNep for various days were also analyzed.Page 3 of 19For EZH2 gene knockdown studies, the cells were transfected with virus constructed of different EZH2 shorthairpin RNA (shRNA) and negative control shRNA bymixed with polybrene (final concentration of 10ug/ml)for 48 h. The transfection mixture was then replacedby complete α-MEM added with puromycin for 24 h,and BMMs were then seeded into a 96-well plate at thedensity of 1 104 cells/well with complete α-MEM. Theculture medium was then replaced every 2 days untilthe matured osteoclasts could be observed in the plate.For Foxc1 inhibition experiment, the cells were separately transfected with negative control siRNA, negativecontrol siRNA with DZNep, and Foxc1 small interferingRNA (siRNA) with DZNep and by mixed with transfection reagent and osteoclast culture medium. The transfection mixture was then replaced every 2 days until thematured osteoclasts could be observed in the plate. Theplate was then washed by PBS 1 times and fixed with 4%paraformaldehyde for 15 min. TRAP staining was performed immediately without light at 37 C for 30 min. Anoptical microscope (Olympus, Tokyo, Japan) was used toimage the photographs, and the TRAP-positive cells withmore than three nuclei were considered osteoclasts andwere quantified the area and number using the ImageJsoftware (NIH, Bethesda, MD, USA).Bone resorption assayTo explore the effect of DZNep on the function of osteoclast, BMMs on complete α-MEM were seeded at thedensity of 1 104 cells/well, in triplicate, on the OsteoAssay Surface plates (Corning, NY, USA). After thirtysix hours, the culture medium was replaced by completeα-MEM, with RANKL (50 ng/mL), and various dosesof DZNep (control, 6.25, 12.5, and 25 nM). For EZH2gene knockdown studies, the cells were transfected withvirus constructed of different EZH2 short hairpin RNA(shRNA) and negative control shRNA by mixed withpolybrene (final concentration of 10ug/ml) for 48 h. Thetransfection mixture was then replaced by completeα-MEM added with puromycin for 24 h, and the survivedBMMs were seeded into a 96-well plate, in triplicate, ata density of 1 104 cells/well. The medium needs to beconstantly replaced until the osteoclast had matured for1 day. For Foxc1 inhibition experiment, the cells wereseparately transfected with negative control siRNA,negative control siRNA with DZNep, and Foxc1 smallinterfering RNA (siRNA) with DZNep and by mixedwith transfection reagent and osteoclast culture medium.The transfection mixture was then replaced every 2 daysuntil the osteoclast had been observed for 2 days. The5% sodium hypochlorite was used to wash the wells for3 min to remove the cells. The total area of resorption

Cao et al. Stem Cell Research & Therapy(2021) 12:605was photographed and then calculated by ImageJ software (NIH, Bethesda, MD, USA).Immunofluorescence of podosome actin beltBMMs in complete α-MEM with or without 25 nMDZNep were seeded in 6-well plates at a density of1 105 cells/well and cultured for 6 days. Once thematured osteoclast could be observed, the plate waswashed by PBS and then fixed with 4% paraformaldehyde. The 10 min 0.2% Triton X-PBS was then performedfor permeabilizing. After three times washing by PBS,FITC-labeled phalloidin was added for 1 h in darknessto bounds with the cells F-actin ring. The nuclei werestained at RT for 5 min without light by 4′, 6-diamidino2-phenylindole (DAPI). After the final PBS wash, theperimeter of the actin ring could be observed by the fluorescence microscope (Leica) and analyzed using ImageJsoftware (NIH, Bethesda, MD, USA).Cultivation, differentiation, and mineralizationof osteoblastsPrimary BMSC were obtained from the whole bone marrow of male 4-week-old C57BL/6 mice. In brief, cellswere extracted from the tibiae and femurs bone marrow and suspended in α-MEM supple with 10% FBS, and1% penicillin/streptomycin. Then, the cell cultures werekept in a humid environment at 37 C with 5% CO2 untilthey reached 90% confluence. For further identifying theeffect of DZNep on osteogenesis, bone marrow-derivedmouse mesenchymal stem cells were firstly seeded intothe 48-well plates for one day at a density of 5 104cells/well. Then the medium was changed by low-glucose DMEM with 15% FBS, 5 mM β-glycerophosphate,50 µg/mL ascorbic acid, 10-7 mM dexamethasone, andvarious concentrations of DZNep (0, 12.5, 25, 50, and100 nM). The medium was changed twice a week. ForEZH2 gene knockdown studies, the cells were transfected with virus constructed of different EZH2 shorthairpin RNA (shRNA) and negative control shRNA bymixed with polybrene (final concentration of 10ug/ml)for 48 h. The transfection mixture was then replaced bycomplete α-MEM added with puromycin for 24 h, andBMSCs were then seeded into a 48-well plate at the density of 5 104 cells/well with complete α-MEM for 24 h.The culture medium was then replaced by low-glucoseDMEM with 15% FBS, 5 mM β-glycerophosphate, 50 µg/mL ascorbic acid, 10-7 mM dexamethasone. The mediumwas changed twice a week. For Wnt4 inhibition experiment, the cells were separately transfected with negativecontrol siRNA, negative control siRNA with DZNep, andWnt4 small interfering RNA (siRNA) with DZNep and bymixed with transfection reagent and osteoblast culturemedium. The transfection mixture was then replacedPage 4 of 19every 3 days. On day 7 and day 28, cells were stainedby using BCIP/NBT kit (Beyotime, Shanghai, China) toobserve the percent of alkaline phosphatase (ALP)-positive cells and 1% Alizarin red S solution (Solarbio, Beijing, China) to visualize the extracellular matrix calciumdeposition.Quantitative real‑time PCR analysisBMMs were cultured in 6-well plates at the densityof 2 105 cells/well in α-MEM supplemented withM-CSF (30 ng/ml), RANKL (50 ng/ml). The cells werethen treated with or without 25 nM DZNep for 0, 1, 3,5 days, or with various concentrations of DZNEP (0, 6.25,25 nM) for 5 days, or with negative control siRNA, negative control siRNA with DZNep, and Foxc1 small interfering RNA (siRNA) with DZNep for 5 days. Once theculture completed, the Axygen RNA Miniprep Kit (Axygen, Union City, CA, USA) was used to extract the totalRNA. BMSC were seeded in a 6-well plate at a numberof 4 105 cells/well and treated with 0, 25, 50, 100 nMof DZNep combined with osteoblast-cultured mediumfor 24 h, or with or without 100 nM DZNep combinedwith osteoblast-cultured medium for 0 and 7 days, orwith negative control siRNA, negative control siRNAwith DZNep, and Wnt4 small interfering RNA (siRNA)with DZNep combined with osteoblast-cultured mediumfor 7 days. Once the culture completed, the Axygen RNAMiniprep Kit (Axygen, Union City, CA, USA) was usedto extract the total RNA. After reverse transcription toobtain cDNA from the RNA template by using the PrimeScript RT reagent Kit was finished. The real-time PCRassay was subsequently performed by using the SYBR Premix Ex Taq II on an ABI 7500 Sequencing Detection System (Applied Biosystems, Foster City, CA). Specifically, total volume of 10 μl liquid consisted by 5 μl ofSYBR Premix Ex Taq II, 1 μl of diluted cDNA, 0.4 μl ofmixed forward and reverse primer, and 3.6 μl of ddH2Owas added into each hole of 384 PCR plate. Conditions of the 40 cycles were: 95 C for 5 s and 60 C for30 s. The melting curves and reverse transcription PCR(RT-PCR) were the methods to detect the specificity ofamplification product. Each target’s quantity, run intriplicate, was normalized to Actin Beta (Actb). MouseActb, CathepsinK (Ctsk), tartrate-resistant acid phosphatase/acid phosphatase (Trap/Acp5), nuclear factor ofactivated T-cells c1 (Nfatc1), AP-1 transcription factorsubunit (c-Fos), Mus musculus ATPase, H transporting, lysosomal V0 subunit D2 (Atp6v0d2), calcitoninreceptor (Calcr), RUNX family transcription factor 2(Runx2), Wnt1, Wnt10a, Wnt4, osteocalcin (Ocn), Wnt6,osteopontin (Opn), and collagen type 1 alpha 1 (Col1a1),primer sequences are listed in Table 1.

Cao et al. Stem Cell Research & Therapy(2021) 12:605Table 1 Primer pairs sequences against osteoclast gene used inqPCRMouse gene Forward 5′ 3′Reverse 5′ 3′Trap/Acp5CAA AGA GAT CGC CAG AAC CGGAG ACG T TG CCA AGG TGA TCCtskCTT CCA ATA CGT GCA GCA GATCT TCA GGG C TT TCT CGT TCAtp6v0d2GCA GAG C TG TAC T TC AAT GTGG TAG TCC GTG GTC TGG AGA TGCalcrTCT GCG T TC C TG AGA ACA CCAAG GCG C TC TAA TGG CAC T TNfatc1TGC TCC TCC TCC TGC TGC TC GCA GAA GGT GGA GGT GCA GCc-FosCCA GTC AAG AGC ATC AGC AAAAG TAG TGC AGC CCG GAG TARunx2TGG CCG GGA ATG ATG AGA ACGGA TGA GGA ATG CGC CCT AAAlplGGG CAT TGT GAC TAC CAC TCG CCT C TG GTG GCA TCT CGT TAT OcnGCG C TC TGT C TC TCT GAC C T TTT GTA GGC GGT C TT CAA GCOpnTTT GTA GGC GGT C TT CAA GCGTG AGA T TC GTC AGA T TC ATCCG Col1a1TAA GGG TCC CCA ATG GTG AGA GGG TCC C TC GAC TCC TAC ATWnt1ATA GCC TCC TCC ACG AAC C T GAT GAA CGC TGT T TC TCG GCWnt4GAG CAA C TG GCT GTA CCT GGGGA ACT GGT ACT GGC ACT CCWnt6GGT CAC TCA AGC C TG T TC CACCG AAG TCC ACA TCG TCT CCWnt10aCTG AAC ACC CGG CCA TAC TTC CTG AAC ACC CGG CCA TAC TTC Foxc1CAC TCG GTG CGG GAA ATG TGTG CGG TAC AGA GAC TGA CTG ActbACA GCA GTT GGT TGG AGC AAACG CGA CCA TCC TCC TCT TAWestern blotting analysisTo explore the variation of the proteins on long-termactivated signaling, BMMs were cultured in α-MEM containing M-CSF (30 ng/ml) and RANKL (50 ng/ml) at adensity of 2 105 cells/well on 6-well plates. After 24 h,the BMMs were treated with or without 25 nM DZNepfor 0, 1, 3, and 5 days, and total protein was obtained,respectively, on these specific time points. To determinethe effect of DZNep on the protein of short-time activated phosphorylated, RAW264.7 cells were cultured inα-MEM at a density of 5 105 cells/well on 6-well plates.One day later, the cells were treated with the serum-freeα-MEM in the presence or absence of DZNEP for 3 handthen stimulated with 50 ng/mL RANKL for 0, 10, 20, 30,and 60 min. To determine the efficiency of EZH2 shRna,BMMs were cultured in α-MEM at a density of 5 105cells/well on 6-well plates. One day later, the cells weretreated with negative control shRNA or different EZH2shRNA for 2 days. To determine the efficiency of Foxc1Page 5 of 19siRna, BMMs were cultured in α-MEM at a density of5 105 cells/well on 6-well plates. One day later, the cellswere treated with negative control siRNA or differentFoxc1 siRNA for 2 days. To identify the role of DZNepon the EZH2 and H3K27me2 in osteoblasts, BMSC cellswere cultured with α-MEM supplemented 10% FBS at adensity of 5 105 cells/well on 6-well plates. After oneday, different doses of DZNEP were treated on the cellsfor 4 h; then, osteoblast -cultured medium was addedfor another 24 h. To determine the efficiency of EZH2shRna, BMSC were cultured in α-MEM at a density of5 105 cells/well on 6-well plates. One day later, the cellswere treated with negative control shRNA or differentEZH2 shRNA for 2 days. To determine the efficiency ofWnt4 siRna, BMSC were cultured in α-MEM at a density of 5 105 cells/well on 6-well plates. One day later,the cells were treated with negative control siRNA or different Wnt4 siRNA for 2 days. Afterward, the cells werewashed by the phosphate-buffered saline (PBS), and totalprotein of the cells was extracted by using lysate consisting of protease, phosphorylase inhibitor cocktail (SigmaAldrich), and radioimmunoprecipitation assay (RIPA)lysis buffer (Beyotime, Shanghai, China). After centrifuged at 13,000 g for 15 min, the bicinchoninic acid(BCA) assay was then used to detect the concentrationsof the protein in the supernatant. Finally, the supernatantproteins were diluted by 5 SDS-sample loading bufferand separated by 4–20% SDS-PAGE before transferredto nitrocellulose filter membranes (GE Healthcare LifeSciences, Pittsburgh, PA, USA). The 5% skim milk dissolved in 1 TBST (Tris-buffered saline with Tween 20)was used to block the membranes at room temperaturefor 1 h. And the primary antibodies (β-actin, 1:1000;LaminB, 1:1000; ERK, 1:1000; p-ERK, 1:1000; JNK,1:1000; p-JNK, 1:1000; p38, 1:1000; p-p38, 1:1000; IKKβ,1:1000; p-IKKα/β, 1:500; IκBα, 1:1000; p- IκBα, 1:1000;P65, 1:1000; p-P65, 1:500; NFATc1, 1:500; c-FOS, 1:1000;EZH2, 1:1000; H3K27me3, 1:1000; GSK-3β, 1:1000; andp-GSK-3β, 1:1000; Foxc1, 1:1000; Wnt4, 1:1000) werethen incubated with the membranes together overnightat 4 C. After TBST wash for 3 times, the secondary fluorescence antibodies were incubated without light for 1 hat room temperature and the reactivity was visualizedby using Odyssey V3.0 image scanning (Li-COR. Inc.,Lincoln, NE, USA). The unseparated Western blottingimages are included in the supplementary material.Osteoblasts and osteoclasts co‑cultivationFor the osteoblasts and osteoclasts co-culture system, calvarial OB isolated from day 1–3 pups werecultured for 24 h. Then the calvarial OB with complete α-MEM (α-MEM supplemented with 10% FBS,1% penicillin/streptomycin) were seeded into 24-well

Cao et al. Stem Cell Research & Therapy(2021) 12:605plates at a density of 2 104 cells/well. After 3 days ofcultivation, the medium was replaced by the completeα-MEM containing 50 ng/ml ascorbic acid and 5 mMβ-glycerophosphate. On the same day, bone marrowmonocytes were isolated from bone marrow and cultured in 10-cm dish. Another 3 days later, BMMs wereseeded at a density of 4 104 cells/well into the 24-wellplate containing the calvarial OB, and the α-MEM supplemented with 15% FBS, 1% penicillin/streptomycin,1, 25-dihydroxy vitamin D3 (10 nM; CSN), and PGE2(1 μM; CSN). After changed every 3 days for 2 times ofthe culture medium, the cells were then fixed and stainedto observe the percent of the positive ALP staining areausing a kit according to the manufacturer’s instructions(Beyotime, Shanghai, China). The percent of area wasthen quantified using ImageJ.Tibia cortical and trabecular bone defect mice modelThis study was totally carried out in terms of the guidelines for the Ethical Conduct in the Care and Use ofNonhuman Animals in Research by the American Psychological Association. Moreover, the Animal Care andExperiment Committee of Shanghai Jiao Tong UniversitySchool of Medicine approved the animal experiment ofthis research.The tibia cortical and trabecular bone defect micemodel was established to determine the bone healingeffect of DZNep in vivo. Firstly, all mice (twelve 8-weekold C57/BL6 male mice (approximate weight 20 2 g))were anesthetized by 2% isoflurane, the right tibia wasexposed, and muscles were dissociated. Then a pilothole was first made with a burr drill point in the proximal tibia, but away from the growth plate to some extent.Then the hole was enlarged with a 1 mm reamer. Afterward, the mice were average divided into two groups:(1) DZNep group (injection with 100 μg/kg DZNep);(2) Sham group (injection with 1 PBS). Two days afteroperation, 1 PBS for sham group and DZNep for treatment group were intraperitoneally injected every twodays during next two week. All mice were finally euthanized after experiment, and whole tibia bones were separated. After fixed in 4% paraformaldehyde for 24 h, the75% ethanol was then used to soak the specimens for further histological and radiographic analysis.Micro‑computed tomography scanningHigh-resolution micro-CT (μCT-100, SCANCO MedicalAG, Switzerland) was used to perform the micro-computed tomography (CT) scanning. Scans were conductedin 75% ethanol and used an X-ray intensity of 200 uA, anX-ray tube potential of 70 kVp, and an integration time of300 ms. For analysis bone mass of defect section, a regionof defect section trabecular bone was contoured, startingPage 6 of 19from the first defect section level away from proximalend of the distal femoral growth plate. Femoral trabecular bone was thresholded at 211 per mille. The indicatorsof bone microstructure (bone volume/tissue volume (BV/TV), trabecular thickness (Tb.Th), trabecula number(Tb.N), trabecula separation (Tb.Sp), and bone surface/bone volume (BS/BV)) were measured by software (Version: 6.5–3, SCANCO Medical AG, Switzerland) throughevaluating and analyzing the three-dimensional region ofinterest (ROI).Histological analysisAfter fixing in 4% paraformaldehyde for 48 h, the 10%EDTA was used to decalcify the fixed tibias for 21 days.Then the tissues were processed by dehydrating in different concentrations of ethanol, infiltrating with xylene,and embedding with paraffin. Hematoxylin and eosin(H&E) and TRAP staining were then performed on theseprepared histological sections. As for immunofluorescentstaining, slices were processed in a sequence of deparaffinization, hydration, antigen retrieval, permeabilization, blocking, primary antibody incubation (OCN, USA,Affinity; dilution 1:100), secondary antibody incubation, and nuclear staining finally. Immunohistochemical(IHC) staining was accomplished with antibodies againstOCN (USA, Affinity; dilution 1:100) [29]. The specimenswere observed and photographed under the high-qualitymicroscope (Leica DM4000B). Each sample’s number ofosteoclasts and TRAP-positive multinucleated osteoclasts per field (Oc.S/BS) were calculated.Statistical analysisThe results were analyzed by using Prism 8 (GraphPadSoftware Inc, San Diego, CA, USA). The data were uniformly presented in the form of median and interquartile range. The comparisons between experimental andcontrol groups were made by the 2-tailed, unpaired Student’s t test, one-way ANOVA with Tukey’s post hoc test.Statistical significance was determined to be at *p 0.05;**p 0.01; ***p 0.001; ****p 0.0001.ResultsDZNep promoted RANKL‑induced osteoclastogenesisin vitroThe cytotoxicity effect of DZNep (Fig. 1A) on primaryBMMs was investigated. As shown in Fig. 1B, no cytotoxicity was found at the concentration of 25 nM. The inhibitory concentration (IC50) value of DZNep was examinedto be 325.7 nM at 96 h (Fig. 1C). To further explore therole of DZNep on osteoclastogenesis in vitro, BMMswere cultured with RANKL, M-CSF and DZNep at concentrations less than or equal to 25 nM. A little numberof multinucleated trap-positive osteoclasts were observed

Cao et al. Stem Cell Research & Therapy(2021) 12:605Page 7 of 19Fig. 1 DZNep promoted RANKL-induced osteoclastogenesis without cytotoxic effects in vitro. A The structure of DZNep. B Cell viability ofDZNep-treated BMMs was tested by CCK-8 at 24, 48, and 96 h. C IC50 value obtained for the activity of DZNep against BMMs. D BMMs were treatedwith different concentrations of DZNep and stimulated by M-CSF (30 ng/ml) and RANKL (50 ng/ml) for 5 days. Then the cells were f

Bone, a dynamic tissue, constantly undergoes the pro - cedure of new bone formation and old bone elimination [4]. Under physiological conditions, this kind of process is inseparably balanced and coordinated by the bone-resorbing osteoclasts and bone-forming osteoblasts [5, 6]. Once pathological damage such as defect and frac-