Deficiency of Macf1 in osterix expressing cells decreases bone formation by Bmp2/Smad/Runx2 pathway

Abstract Microtubule actin cross‐linking factor 1 (Macf1) is a spectraplakin family member known to regulate cytoskeletal dynamics, cell migration, neuronal growth and cell signal transduction. We previously demonstrated that knockdown of Macf1 inhibited the differentiation of MC3T3‐E1 cell line. However, whether Macf1 could regulate bone formation in vivo is unclear. To study the function and mechanism of Macf1 in bone formation and osteogenic differentiation, we established osteoblast‐specific Osterix (Osx) promoter‐driven Macf1 conditional knockout mice (Macf1f/fOsx‐Cre). The Macf1f/fOsx‐Cre mice displayed delayed ossification and decreased bone mass. Morphological and mechanical studies showed deteriorated trabecular microarchitecture and impaired biomechanical strength of femur in Macf1f/fOsx‐Cre mice. In addition, the differentiation of primary osteoblasts isolated from calvaria was inhibited in Macf1f/fOsx‐Cre mice. Deficiency of Macf1 in primary osteoblasts inhibited the expression of osteogenic marker genes (Col1, Runx2 and Alp) and the number of mineralized nodules. Furthermore, deficiency of Macf1 attenuated Bmp2/Smad/Runx2 signalling in primary osteoblasts of Macf1f/fOsx‐Cre mice. Together, these results indicated that Macf1 plays a significant role in bone formation and osteoblast differentiation by regulating Bmp2/Smad/Runx2 pathway, suggesting that Macf1 might be a therapeutic target for bone disease.

In previous studies, Macf1 has been found to participate in the regulation of osteoblast differentiation-associated Wnt/β-catenin signalling pathway. 5,6,14,15 In our previous studies, it was demonstrated that Macf1 can regulate the proliferation, cell cycle progression and differentiation of MC3T3-E1 osteoblastic cell line. [16][17][18][19] However, it remains unknown whether Macf1 could regulate bone formation in vivo.
The bone morphogenetic protein 2 (Bmp2) signalling pathway is a critical regulator of osteogenesis. 20,21 Bmp2 binds to its receptors to induce phosphorylation of Smad1/5/9. Activated Smads can form hetero complexes with Smad4, and then, the complexes are translocated into nucleus to regulate target genes such as runt-related transcription factor 2 (Runx2) and osterix (Osx). 22 Runx2 is a master transcription factor necessary for osteoblast differentiation, which can regulate the expression of osteoblast-specific genes including alkaline phosphatase (Alp), collagen type I (Col1) and osteocalcin (Ocn). 23 It has been reported that Wnt/β-catenin pathway can regulate the activation of BMP2 transcription in osteoblasts. 5,24 Thus, we hypothesized that Macf1 could regulate osteoblast differentiation by modulating Bmp2 pathway.
To investigate the role of Macf1 in bone formation and osteoblast differentiation, we generated a mice model in which Macf1 was specifically deleted in osteoblasts by Cre-loxP recombination system. Here, we showed that deficiency of Macf1 decreased bone mass, deteriorated bone microarchitecture and impaired bone strength. In addition, we found that knockout of Macf1 inhibited the differentiation of primary osteoblasts through Bmp2/Smad/Runx2 pathway. Our studies revealed novel insights into the function and mechanism of Macf1 in bone formation. Moreover, we provided a new mice model for in vivo function research of Macf1 and targeted therapy research of osteoporosis.

| Generation of Macf1 f/f Osx-Cre mice
Osterix-Cre (Osx-Cre) and Macf1 flox/flox (Macf1 f/f ) C57BL/6 mouse lines used in the current study have been previously described. 6,25 Osx-Cre mice were crossed with Macf1 f/f mice, and their progeny were bred to obtain osteoblast-specific conditional knockout mice (Macf1 f/f Osx-Cre). Macf1 f/f mice were used as control.
The genotypes were determined by PCR amplification of genomic DNA isolated from the toes or tails of newborn mice. PCR was conducted in an BIO-GENER GE4852T thermocycler (BIO-GENER) with an initial denaturation at 95°C for 5 minutes; then 35

| Primary osteoblasts isolation and osteogenic differentiation
Primary osteoblasts were isolated from calvaria of mice on postnatal day 1 as previously described. 26 Briefly, calvaria was dissected and sequentially digested by digestion solution containing 0.1% type II collagenase (Invitrogen, CA, USA) and 0.2% neutral protease (Kehao) in hanks' balanced salt solution for 15 minutes at 37°C. The cells isolated from the first and second digestions were discarded, and cells obtained from later four digestions were cultured in α-MEM (Gibco) supplemented with 10% FBS (Corning), 1% L-glutamine (Sigma), 100 U/mL penicillin (Amresco) and 100 μg/mL streptomycin (Amresco).

Sequence
Product size TA B L E 1 Primer sequences for genotyping mL recombinant human BMP2 (rhBMP2, PeproTech) was added to osteogenic medium when needed. Then, cells were subjected to Alkaline phosphatase staining at 7 days after differentiation and Alizarin red S staining at 14 days after differentiation.

| Alkaline phosphatase staining
Alkaline phosphatase (ALP) staining was performed with a BCIP/NBT ALP colour development kit (Beyotime). Briefly, cells were rinsed 3 times with PBS and fixed with 4% PFA. Then, the osteoblasts were incubated with 500 μL/well of BCIP/NBT substrate for about 2 hours in the dark. After washing with PBS, the ALP active cells were imaged.

| Western blot analysis
Tibia, liver, spleen and heart of the male mice were immediately separated and pulverized in a cooled mortar with liquid nitrogen.
The bone marrow in tibia was flushed out before pulverization. and visualized by a T5200 Multi chemiluminescence detection system (Tanon). Densitometric quantification of the bands was performed with Image J analysis software (Image J).

| Quantitative real-time PCR (qPCR)
Total RNA was extracted from primary osteoblasts or pulverized tissues by TRIzol reagent (Invitrogen) according to the manufacturer's instructions. 1 μg of total RNA was reverse-transcribed into cDNA by PrimeScript RT reagen Kit (TaKaRa). Real-time PCR assay for mRNA detection was performed by SYBR qPCR Master (TaKaRa) using fluorescence quantitative detection system (Line-Gene 9600, BIOER) with an initial denaturation at 95°C for 30 seconds, followed by 45 cycles at 95°C for 10 seconds, 60°C for 30 seconds and 72°C for 5 seconds. Primers used for qPCR were listed in Table 2. Gapdh was used as control for mRNA analysis.

| Dual energy X-ray absorptiometry (DXA)
After anesthetized with 15 μg/mL pentobarbital sodium, mice were placed on the specimen tray of DXA body composition analysis system (InAlyzer, Medikors) in a prone or side position for whole body TA B L E 2 Primer sequences for qPCR scanning. Radiographic images and related parameters of different bone regions were obtained using InAlyzer Dual X-ray Digital Imaging Software (InAlyzer, Medikors).

| Micro-Computed Tomography (micro-CT)
Micro-Computed Tomography analysis was performed as previously described. 27 Briefly, the distal femurs were scanned ex vivo by a micro-CT system (viva CT40, Scanco Medical) with a voxel size of

| Bone mechanical properties
The biomechanical strength of femurs was analysed by three-point bending test performed on an Instron 5943 universal test machine (Instron). The span between the lower supports was set to 10 mm.
The upper contact point was aligned at the midpoint of the lower supports. The upper actuator moved at a rate of 1.5 mm/min till the specimen was failed, and the load-displacement curves were received. The inner and outer diameters of loaded bones were measured by a 3D digital microscope (Hirox KH-8700). Other parameters including fracture energy, young's modulus, peak bending stress and strain were determined using standard algorithms as previously described. 28

| Statistical analyses
Statistical analysis was performed using Graph Pad PRISM 5.0.
All results are expressed as mean ± SD. Unpaired Student's t tests were used to compare data between two Macf1 f/f and Macf1 f/f Osx-Cre mice groups. For all experiments, significance was defined as *P < .05, **P < .01 and ***P < .001.

| Generation of Macf11 conditional knockout mice
To

| Deficiency of Macf1 delayed bone ossification and decreased bone mass
To determine the effects of Macf1 deficiency on bone ossifica-

| Deficiency of Macf1 impaired bone biomechanical properties
The effect of bone quality on biomechanical properties is determined by the material composition and microarchitecture of bone. 29 Based on the above findings, the biomechanical strength of the femurs from 3-month-old Macf1 f/f and Macf1 f/f Osx-Cre mice was analysed by three-point bending test. As shown in Figure 3F, the   Figure 4B,4). In addition, IHC assay showed that the expression of Runx2 and Ocn was decreased in the femur of 3-month-old Macf1 f/ f Osx-Cre mice ( Figure 4D). Taken together, these data suggested that osteoblast-specific deletion of Macf1 inhibited the differentiation of primary osteoblasts. was rescued by exogenous rhBMP2 ( Figure 5C). ALP and ARS staining results showed that the differentiation capability was significantly recovered in primary osteoblasts from Macf1 f/f Osx-Cre mice by rhBMP2 ( Figure 5D). Expressions of osteogenic marker genes in osteoblasts were also recovered after rhBMP2 treatment ( Figure 5E). These results demonstrated that deficiency of Macf1 inhibited osteogenic differentiation of primary osteoblasts via Bmp2/Smad/Runx2 pathway.

| D ISCUSS I ON
In the present study, we constructed an osteoblast-specific Osterix Macf1 f/f Osx-Cre without rhBMP2 treatment that osteoblasts and adipocytes in bone marrow share a common progenitor, and the dynamic balance between them can influence the density and function of bone. 35 In this study, histological staining of Macf1 deficiency femur displayed an increased accumulation of adipocyte ghosts in situ ( Figure 3E). This phenotype is similar to the observations following the β-catenin and Cbfβ deletion in Osx-expressing cells. 36,37 It would be interesting to examine the function and mechanism of Macf1 on bone marrow adiposity in the future.
Osteoblasts are uniquely responsible for bone formation. 38 Our previous studies showed that knockdown of Macf1 inhibited the differentiation of MC3T3-E1 osteoblastic cell line in vitro. 17  and Macf1 f/f Osx-Cre mice, the down-regulation of the Smads and Runx2 in Macf1-delected osteoblasts was significantly recovered.
In addition, the differentiation ability of Macf1 knockout osteoblasts was rescued accordingly. Together with the present results, we concluded that Macf1 could regulate osteoblast differentiation through Bmp2/Smad/Runx2 signalling pathway.
In summary, we successfully generated an osteoblast-specific Macf1 conditional knockout mice model using Cre-loxP system. Our data demonstrated that deletion of Macf1 decreased bone mass, deteriorated bone microarchitecture and impaired bone strength.
Macf1 could affect the differentiation and mineralization of osteoblasts through Bmp2/Smad/Runx2 signalling pathway. Our study not only reveals a novel role and mechanism of Macf1 in bone formation, but also provides a mice model to further study the functions of Macf1 in vivo.

ACK N OWLED G EM ENTS
This work was supported by grants from National Natural

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflicts of interest to disclose.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.