EAPB02303

Mechanical unloading reduces microtubule actin crosslinking factor 1 expression to inhibit β-catenin signaling and osteoblast proliferation

Chong Yin1,2,3#, Yan Zhang1,2,3#, Lifang Hu1,2,3, Ye Tian1,2,3, Zhihao Chen1,2,3, Dijie Li1,2,3, Fan Zhao1,2,3, Peihong Su1,2,3, Xiaoli Ma1,2, Ge Zhang2,3, Zhiping Miao1,2,3, Liping Wang4,*, Airong Qian1,2,3,* , Cory J. Xian4

ABSTRACT

Mechanical unloading was considered a major threat to bone homeostasis, and has been shown to decrease osteoblast proliferation although the underlying mechanism is unclear. Microtubule actin crosslinking factor 1 (MACF1) is a cytoskeletal protein that regulates cellular processes and Wnt/ β-catenin pathway, an essential signaling pathway for osteoblasts. However, the relationship between MACF1 expression and mechanical unloading, and the function and the associated mechanisms of MACF1 in regulating osteoblast proliferation are unclear. This study investigated effects of mechanical unloading on MACF1 expression levels in cultured MC3T3-E1 osteoblastic cells and in femurs of mice with hind limb unloading; and it also examined the role and potential action mechanisms of MACF1 in osteoblast proliferation in MACF1-knockdown, overexpressed or control MC3T3-E1 cells treated with or without the mechanical unloading condition. Results showed that the mechanical unloading condition inhibited osteoblast proliferation and MACF1 expression in MC3T3-E1 osteoblastic cells and mouse femurs. MACF1 knockdown decreased osteoblast proliferation, while MACF1 overexpression increased it. The inhibitory effect of mechanical unloading on osteoblast proliferation also changed with MACF1 expression levels. Furthermore, MACF1 was found to enhance β-catenin expression and activity, and mechanical unloading decreased β-catenin expression through MACF1. Moreover, β-catenin was found an important regulator of osteoblast proliferation, as its preservation by treatment with its agonist lithium attenuated the inhibitory effects of MACF1-knockdown or mechanical unloading on osteoblast proliferation. Taken together, mechanical unloading decreases MACF1 expression, and MACF1 up-regulates osteoblast proliferation through enhancing β-catenin signaling. This study has thus provided a mechanism for mechanical unloading-induced inhibited osteoblast proliferation. This article is protected by copyright. All rights reserved

Keywords: Mechanical unloading, MACF1, Osteoblast proliferation, β-catenin

1. Introduction

Mechanical stimulation plays an essential role in regulating bone remodeling and maintaining bone homeostasis (Alzahrani et al., 2014; Imbert et al., 2014). The mechanical unloading due to extended bed rest or immobilization would result in significant bone loss (Gaudio et al., 2010; Oppl et al., 2014). Osteoblasts are responsible for bone formation and play a key role in regulating bone remodeling (Papachroni et al., 2009). Both numerical and functional decreases of osteoblasts will cause bone loss. Previous studies have shown that mechanical unloading treatment decreased osteoblast proliferation rate (Luo et al., 2013; Morey & Baylink, 1978; Wronski & Morey, 1983); however, mechanism for this still remains unclear. Cytoskeletal proteins have been proved sensitive to mechanical stimulation and were known as the mechanical sensors in cells (Ingber, 1997). Microtubule actin crosslinking factor 1 (MACF1), also named ACF7 (actin crosslinking factor 7), is a member of the plakin family (Hu et al., 2016; Leung et al., 2002) . It is a 600 kDa large cytoskeletal protein that crosslinks with actin and microtubules (Karakesisoglou et al., 2000; Kodama et al., 2003; Leung et al., 1999). It is ubiquitously expressed in different tissues (Hu et al., 2016; Suozzi et al., 2012). Its gene mutation or abnormal expression has been shown to correlate with various diseases (Afghani et al., 2017; Goryunov & Liem, 2016; Ka et al., 2014; Liang et al., 2013). As a cytoskeletal protein, MACF1 plays an essential role in controlling actin and microtubule cytoskeletal dynamics (Kodama et al., 2003; Wu et al., 2008; Wu et al., 2011; Yue et al., 2016), and it regulates cardiomyocyte microtubule distribution in response to hemodynamic overload (Fassett et al., 2013). MACF1 may play an essential role in osteoblasts responding to mechanical unloading, since both MACF1 expression and its association with actin and microtubule cytoskeleton were altered when the MC3T3-E1 and MG-63 osteoblastic cells were under the diamagnetic levitation-induced mechanical unloading condition (Qian et al., 2009). Genetically, the knockdown of MACF1 inhibited MC3T3-E1 cell proliferation (Hu et al., 2015). These results imply that MACF1 expression might be correlated with mechanical unloading and involved in regulating osteoblast proliferation. However, the effects of mechanical unloading on MACF1 expression, as well as the functions and action mechanisms of MACF1 regulating osteoblast proliferation are still unknown.
As a cytoskeletal protein, MACF1 has been reported to regulate intracellular transport (Burgo et al., 2012; Chen et al., 2006). Previous studies have demonstrated that MACF1 acts upstream of β-catenin and mediates its nuclear importation (Chen et al., 2006; Hu et al., 2017). In addition, the mechanical unloading condition has been shown to inhibit the nuclear importation of β-catenin (Case et al., 2008; Robinson et al., 2006), which was quite similar with the phenomenon when MACF1 expression was inhibited (Hu et al., 2017). As an essential factor for bone development and bone formation (Kang & Robling, 2014) , β-catenin regulates osteoblast proliferation and osteogenic differentiation (Galindo et al., 2007) . However, whether the β-catenin function and its regulation are linked with MACF1 function remain unclear. In the current study, our hypothesis was that mechanical unloading inhibits osteoblast proliferation by reducing the expression of MACF1, which impedes the nuclear importation of β -catenin. Using in vitro and in vivo models, the current study examined the effect of mechanical unloading on MACF1 expression and investigated the relationship between MACF1 expression, osteoblast proliferation, and β-catenin activity.

2. Materials and methods

2.1 Plasmids, Cell culture, random positioning machine application and hind limb unloading mouse model

MACF1 overexpression plasmid PEGFP-C1A-ACF7 (Wu et al., 2008) was generously provided by Dr. Xiaoyang Wu (The University of Chicago, Chicago, IL). Control plasmid PEGFP-C1A was purchased from Genechem Int. (Shanghai, China). TOPflash reporter plasmid and Renilla pRL-TK plasmid were generously provided by Dr. Pengsheng Zheng (Xi’an Jiaotong University, Xi’an, China). Murine preosteoblast cell line MC3T3-E1 was generously provided by Dr. Hong Zhou (The University of Sydney, Sydney, Australia). The MC3T3-E1 cells were cultured in alpha Modified Eagle’s Medium (α-MEM, Gibco, 11900-024, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS; Biological Industries, 04-001-1A, Kibbutz Beit Haemek, Israel), 1% L-glutamine (Sigma, G8540, St Louis, MO), 1% penicillin (Amresco, 0242, Solon, OH) and streptomycin (Amresco, 0382, Solon, OH). Cell cultures were maintained at a humidified, 37°C, 5% CO2 incubator (Thermo Fisher Scientific, Waltham, MA).
To confirm the inhibitory effects of mechanical unloading on osteoblast proliferation, we used the Random Positioning Machine (RPM; the Center for Space Science and Applied Research of Chinese Academy of Sciences, Beijing, China), which was designed to simulate the effects of mechanical unloading condition on plant and animal cells, as described previously (Chen et al., 2016; Hu et al., 2013; Pardo et al., 2005; Qian et al., 2012; Van Loon, 2007). Cell culture vessel was fixed on the inner frame and the RPM was placed inside a 37°C incubator (Thermo Fisher Scientific). For mechanical unloading studies, MC3T3-E1 cells (or MACF1-knockdown and MACF1-overexpression MC3T3-E1 cells) were seeded at a density of 5×10 3 cells/cm2 on glass coverslips in 90 mm dish and incubated at 37°C. RPM culture flasks filled with growth medium (air bubbles avoided) were tightly capped and mounted in the inner frame of the RPM. The machine was operated in a random mode of speeds (0-10 revolutions per minute) and directions including both inner and outer frames (Van Loon, 2007). Cells of the static control group were cultured in the same 37°C incubator but without rotation.
A hind limb unloaded mouse model was adopted to simulate the mechanical unloading condition. Sixteen 4-month-old male C57BL/6 mice (22.6 ± 0.9g) were purchased from the Laboratory Animal Center of the Fourth Military Medical University (Xi’an, China). Mice were randomly divided into two groups of eight mice each: HLU (hind limb unloaded) and control. The HLU group was kept for four weeks with hind limbs suspended, while the control group was normally loaded. The hind limb suspension was achieved by tail suspension as previously described (Wronski & Morey, 1983). In brief, mice were kept in standard cages with the suspension position maintained at an angle of about 30°. This maneuver permitted the animals still to have ad libitum access to food and water. The animal’s overall appearance, drinking and eating habits, and tail were checked two times per day. Euthanasia was performed at week 4 by CO2 gas overdose. The left femurs were collected and processed for bone-derived mesenchymal stem cells isolation (n=8/group). The right femurs were separated into two parts for Western blot (n=4/group) and RT-PCR analyses (n=4/group) as described below. All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals: Eighth Edition, and all experimental procedures were approved by the Institutional Experimental Animal Committee of Northwestern Polytechnical University, Xi’an, China. For all procedures involving animals, all efforts were made to reduce the number of the mice used and their suffering.

2.2 Isolation of bone-derived mesenchymal stem cells (MSCs )

To investigate the effect of mechanical unloading condition on MACF1 expression in osteogenic cells in vivo , bone-derived MSCs were isolated from HLU mouse femurs using a previously published protocol (Zhu et al., 2010). Briefly, femoral bones were harvested and attached soft tissues were gently removed. A syringe needle was inserted into the bone marrow cavity and bone marrow was removed by flushing several times with phosphate buffered saline (PBS). Then, the diaphysis of femurs was cut into small pieces, and cultured in a 60mm plate for 3 days. The bone pieces were transferred into a new plate for the 1st-passage cells. Cells were cultured for another 7 days. Third passage cells were used for characterization and experiments. The osteogenic characteristic of bone-derived MSCs were assessed by osteoblast differentiation assays as described (Zhu et al., 2010).

2.3 Vi-Cell XR cell viability and cell number analyses

Cell viability and numbers were assessed using Vi-Cell XR cell viability analyzer (Beckman Coulter, Brea, CA) as previously described (Louis & Siegel, 2011; Mundra et al., 2012). This assay measures the cell number and cell viability by applying trypan blue dye-exclusion staining combined with image-based data analysis. For normal and transfected MC3T3-E1 cells, cells were plated into 6-well plates in triplicate at a density of 6×104 /well and cultured for 48h. For RPM-treated cells, cells were seeded at a density of 5×10 3 cells/cm2 on glass coverslips in 90mm dish and then treated by RPM for 24h and 48h. Then, the cells were harvested by digestion with trypsin (Life Technologies, 25300-054, Carlsbad, CA) and resuspended in 0.45ml PBS for automatic cell counting with Vi-Cell analyzer. Viable cell numbers were determined.

2.4 5-Ethynyl-2′-deoxyuridine (EdU) labeling cell proliferation assay

For assessing treatment effects on cell proliferation, EdU labeling was performed using the EdU labeling/detection kit (RiboBio, C10310, Guangzhou, China). For normal and transfected MC3T3-E1 cells, cells were cultured in 24-well plates at a density of 1×10 4 cells per well for 48h at 37°C. For RPM-treated cells, cells were seeded on glass coverslips and treated by RPM for 48h, and the glass coverslips were cut into appropriate size and moved to 24-well plates with 1ml growth medium. Then, 50μM of EdU was added to each well and cells were cultured for additional 8h (6h for plasmid transfected cells) at 37°C. The cells were fixed with 4% formaldehyde (Sigma, F1635) for 15min at 37°C and treated with 0.5% Triton X-100 (Sigma, T9284) for 20min at 37°C. After washing with PBS three times, 100μl of 1× Apollo reaction cocktail was added to each well and the cells were incubated for 30min at 37°C. Then, the cells were stained with 100μl of Hoechst 33342 (RiboBio, C10310) for 30min at 37°C. Cells were visualized at room temperature under an inverted fluorescence microscope (Nikon 80i, Tokyo, Japan) linked to a charge coupled device. All cells were gamma adjusted, merged with Hoechst, counted and analyzed using Image-Pro Plus 6.0 software (Media Cybernetics, Sarasota, FL), the EdU incorporation rate was expressed as the ratio of EdU positive cells to total Hoechst positive cells. All experiments were done in triplicate and three independent repeating experiments were performed (Guo et al., 2011).

2.5 RT-PCR and Western blot analyses

RT-PCR and western blot analyses were used to assess both mRNA and protein levels of MACF1 and components of the β-catenin signaling pathway. Total RNA was extracted from mouse femurs or cultural cells using Trizol reagent (Invitrogen, 15596018, Carlsbad, CA). For mouse femurs, bones were harvested and stored in RNA protector (TaKaRa, 9750, Dalian, China). Soft tissues attached on bone were carefully removed. Bone marrow was flushed as described in 2.2. Diaphysis of femurs was grinded by a pre-cooled mortar with liquid nitrogen and then digested by Trizol reagent. Total RNA of 1μg was used for cDNA synthesis using one step PrimeScript RT reagent kit (TaKaRa, RR037A). Quantitative PCR amplification was performed using the Thermal Cycler C-1000 Touch system (BIO-RAD CFX Manager, Hercules, CA) and SYBR Premix Ex TaqII kit (TaKaRa, RR820A). GAPDH was used as internal control gene. The quantitative PCR reaction conditions included initial denaturation step at 95 oC for 30s, followed by 44 cycles at 95oC for 10s, 60oC for 30s, and 72oC for 5s. Data were calculated using the comparative Ct method (2-ΔΔCt) and expressed as fold change compared with corresponding control. Primers (sequences see Table 1) were synthesized by Sangon Int (Shanghai, China).
The Western blot analysis was performed as previously described (Hu et al., 2015). Protein samples from mouse femurs or cultural cells were extracted. Mouse femurs were harvested, cleared off soft tissues and bone marrow, and stored in 1% protease inhibitor cocktail (Calbiochem, 539134, Darmstadt, Germany). Diaphysis was grinded by mortar with liquid nitrogen and digested by cell lysis buffer (Beyotime, P0013, Haimen, China) with 1% protease inhibitor cocktail on ice for 30mins. As for cultural cells, the cells were washed three times by cold PBS and then digested by cell lysis buffer with 1% protease inhibitor cocktail. Protein concentrations were analysed by BCA protein assay kit (Thermo Fisher Scientific, 23225). Proteins were subjected to SDS-PAGE using 5% stacking gel and 12% separating gel, 120V, 1h (except for MACF1, using 10% stacking gel and 6% separating gel, 200V, 1h) and transferred (400mA, 2h) to nitrocellulose filter membranes (Pall, 66485, Port Washington, NY). Membranes were blocked with 5% skimmed milk (BD Biosciences, 232100, Franklin Lakes, NJ) for 1 hour at room temperature, and then incubated with primary antibodies at 4oC overnight with the following primary antibodies: MACF1 (rabbit pAb, 1:500; Abcam, ab117418, Cambridge, UK), non-phosphorylated β-catenin (rabbit pAb, 1:1000; Cell Signaling Technology, 8814s, Danvers, MA), and GAPDH (mouse mAb, 1:1000; Calbiochem, CB1001). Blots were then incubated with HRP-labeled secondary antibody (CWBIO, (Thermo Fisher Scientific, NCI5080). Protein bands were exposed to X-ray film (Kodak, 6535876, Rochester, NY). GAPDH was adopted as the normal control.

2.6 Immunohistochemical staining

To investigate the effects of mechanical unloading condition on MACF1 expression and distribution, immunohistochemical staining analysis was performed as previously described. Mouse femurs were dissected and fixed in 4% paraformaldehyde (Sigma, P6148), decalcified in 17% ethylene diamine tetraacetic acid (EDTA, Sigma, E9884) for 28 days, and embedded in paraffin (Huayong, Shanghai, China). Sections of 5μm thick were dewaxed, immersed in the distilled water, blocked in 5% goat serum (CWBIO, CW0130) in PBS and then incubated overnight at 4°C with primary antibodies against MACF1 (rabbit pAb, 1:100; Abcam) with non-antibody serum as negative control. Following three washes in PBS, the sections were labeled with HRP-labeled secondary antibody 1h at room temperature and developed for color reaction using diaminobenzidine (DAB, CWBIO, CW2068). MACF1 protein immunostaining intensities were measured by scanning using Aperio AT2 Digital Pathology Scanner (Leica, Wetzlar, Germany) and analyzed by Image-Pro Plus 6.0 software (Jiao et al., 2015) .

2.7 Cell transfection

MACF1-overexpression MC3T3-E1 cells were constructed by transfection of plasmid PEGFP-C1A-ACF7, with PEGFP-C1 transfected cells as control. MC3T3-E1 cells (1×107 per well) were electroplated (1800V, 30ms) with the MACF1 overexpression plasmid PEGFP-C1A-ACF7 or normal plasmid PEGFP-C1A by using Neon Transfection System (Invitrogen). After the electroporation, cells were seeded into a 6-well plate with 10ml α-MEM and cultured for 6h. Adherent cells were then washed by 2ml α-MEM two times and medium was changed to antibiotic-free growth medium (α-MEM with 10% FBS, and 1% L-glutamine). After culture for 48h, medium was changed to the selective growth medium supplemented with 650μg/ml Geneticin (MP Biomedicals, 0215878291, Santa Ana, CA) and cells were cultured for two weeks. Stable MACF1-knockdown MC3T3-E1 cell line was made by transfection of lentivirus vector carrying shRNA targeting murine MACF1 (NM_001199136.1) or its scramble control as described previously (Hu et al., 2015). β-catenin siRNA transfection was preformed to inhibit β-catenin expression. MC3T3-E1 cells were seeded in a 6-well plate at 1×10 5 cells per well and transfected for 6h in serum-free medium with 80nM β-catenin-siRNA (RiboBio, Guangzhou, China) by lipofectamine 2000 (Invitrogen, 11668-030). After transfection, the serum-free medium was replaced by the growth medium. After 48h culture, MC3T3-E1 cells were harvested for analyses by RT-PCR, Western blot, and Vi-Cell XR cell viability analyzer (Hu et al., 2017) . β-catenin-siRNA sequences were: 5’-CAAGCCUUAGUAAACAUAAdTdT-3’ (sense strand) and 3’-dTdTGUUCGGAAUCAUUUGUAUU-5’ (antisense strand).

2.8 Luciferase reporter assays

To analyse the potential function of MACF1 in regulating β-catenin activity in osteoblastic cells, TOPflash plasmid and internal control pRL-TK plasmid were mixed and co-transfected into MACF1-knockdown MC3T3-E1 cells with MACF1-overexpression or its negative control (with plasmid electroporation with plasmid PEGFP-C1A-ACF7 or PEGFP-C1A electroporation using the Neon Transfection System, 2100V, 21ms). A luciferase reporter assay was performed 72h after transfection. Firefly and Renilla luciferase activities were analyzed using the dual-luciferase reagent assay kit (Promega, E1910, Fitchburg, WI). Values for TOPflash luciferase activity were normalized by Renilla activity.
To analyse potential effects of mechanical unloading on regulating β-catenin activity, MC3T3-E1 cells were seeded at a density of 6×103 cells/cm2 on glass coverslips in 90 mm dish at 37°C. Cells were co-transfected with TOPflash plasmid and internal control pRL-TK plasmid by Engreen EntransterTM H4000 Reagent (Engreen, 4000-6, Beijing, China). By 16h after transfection, cells were moved into a flask for RPM treatment, and the luciferase reporter assay was performed after 48h RPM treatment. β-catenin activity after β-catenin siRNA treatment were detected after MC3T3-E1 cells were transfected with β-catenin siRNA as mentioned above. Transfected cells were cultured for 5h and then co-transfected with TOPflash and pRL-TK plasmid by Engreen Entranster TM H4000 Reagent (Engreen). Luciferase reporter assay was performed 48h after the second transfection.

2.9 Statistical analyses

All experiments were independently repeated at least three times with each done in triplicate. Statistical analyses of the data were performed using the GraphPad Prism 6 software (GraphPad Software, La Jolla, CA), and a student t-test was used. All data were reported as the mean ± standard deviation, and P values < 0.05 were considered statistically significant for all comparisons. 3. Results 3.1 Mechanical unloading condition inhibited osteoblast proliferation The potential effect of mechanical unloading on osteoblast proliferation was determined using MC3T3-E1 cells. After 48 hours of random positioning machine (RPM) treatment, Vi-Cell XR cell viability analysis showed that MC3T3-E1 cell numbers was decreased by 31.8%, compared to control (P<0.01, Fig. 1A). Cell proliferation 5-Ethynyl -2′-deoxyuridine (EdU) labeling assays revealed that fewer EdU positive cells were detected in RPM treated cells (Fig. 1B), and EdU positive cell ratio of RPM group was decreased by 14.8% (P<0.001) when compared to control group (Fig. 1C). The results indicated that mechanical unloading inhibited MC3T3-E1 cell proliferation. 3.2 Mechanical-unloading condition suppressed MACF1 expression MACF1 RNA expression in MC3T3-E1 cells was down-regulated by 36% (P<0.01) and 39% (P<0.01) after 24 and 48 hours of RPM treatment, respectively. Moreover, MACF1 protein levels were also decreased as shown by Western blot analyses (Fig. 2A). To examine the mechanical unloading condition on MACF1 expression levels in vivo, the hind limb unloading (HLU) model in C57BL/6 mice was adopted. After 28d HLU treatment, mRNA expression of MACF1 in HLU mouse femur tissue was decreased by 48.5% ( P<0.01) when compared to control, and MACF1 protein levels were also decreased (Fig. 2B). Immunohistochemical (IHC) staining of MACF1 showed a significant decrease in the cortical bone of HLU mice femur compared to control (Fig. 2C, D, supplemental Fig 1). RT-PCR results showed that in bone-derived mesenchymal stem cells (MSCs) isolated from HLU mice, MACF1 expression was decreased by 34.2% when compared to the control mice ( P<0.05, Fig. 2C). 3.3 MACF1 promoted osteoblast proliferation After the transfection of MACF1-overexpression plasmid PEGFP-C1A-ACF7 when compared with the control plasmid PEGFP-C1A, both MACF1 RNA and protein expression levels in MC3T3-E1 cells were increased (Fig. 3A). The transfection also increased the cell numbers by 33.3% ( P<0.05, Fig. 3B), and increased EdU labelling by 24.4% ( P<0.001, Fig. 3C, F). Conversely, in MACF1-knockdown osteoblasts (Hu et al., 2015), cell numbers were decreased by 29.7% ( P<0.01, Fig. 3D) and EdU labelling was decreased by 31.7% ( P<0.001, Fig. 3E, G), when compared to negative control. These results suggest a positive regulatory role of MACF1 for osteoblast proliferation. To investigate potential rescuing effects of MACF1 on osteoblastic cell proliferation, MACF1-knockdown MC3T3-E1 cells and its related negative control cells were then transfected by plasmid PEGFP-C1A-ACF7 or the control plasmid PEGFP-C1A. MACF1 RNA expression level in MACF1-knockdown MC3T3-E1 cells was significantly increased (49.6%, P<0.001) with transfection with the MACF1 overexpression plasmid when compared with the control (Fig. 4A). Similar results were obtained with cell numbers (increased by 73%) (P<0.05, Fig. 4B) and the EdU labelling (increased by 57.8%) (P<0.001, Fig. 4C, D). The increased cell numbers and EdU positive cell ratios proved the rescue effects of MACF1 on osteoblast proliferation. 3.4 Mechanical unloading condition inhibited osteoblast proliferation through MACF1 So far, we have observed that mechanical unloading condition inhibited osteoblast proliferation and suppressed MACF1 expression, and that MACF1 promoted osteoblast proliferation. Next, we investigated whether MACF1 played a role in mechanical unloading condition-induced suppressed osteoblast proliferation. Firstly, MACF1-knockdown MC3T3-E1 cells and their negative control cells underwent RPM treatment and the treatment effects on their MACF1 RNA expression levels were compared via RT-PCR analyses. For negative control cells, MACF1 RNA expression levels were decreased by 46.7% (P<0.001) after RPM treatment. However, in MACF1-knockdown cells, MACF1 expression levels were decreased only by 14.2% upon RPM treatment (P<0.05, Fig. 5A). Secondly, RPM treatment effects on cell proliferation rates were compared. While the proliferation rates of negative control cells were significantly decreased, no differences were found in proliferation rates of MACF1-knockdown cells upon RPM treatment (Fig. 5B, C, F). Thirdly, MACF1 RNA expression levels were found to increase in PEGFP-C1A-ACF7 transfected RPM-treated cells ( P=0.069), but not in RPM-treated cells transfected with PEGFP-C1A control plasmid (Fig. 5D). Lastly, Vi-Cell XR cell viability analysis also showed a significant increase in cell numbers of PEGFP-C1A-ACF7 transfected RPM-treated cells ( P<0.001, Fig. 5E), but a non-significant increase in cell numbers of in RPM-treated cells transfected with PEGFP-C1A control plasmid. These results suggest that mechanical unloading condition regulated osteoblast proliferation through MACF1. 3.5 MACF1 regulated β-catenin expression and activity in osteoblasts Since it has been shown previously that MACF1 acts upstream of β-catenin and mediates its nuclear importation (Chen et al., 2006) and that mechanical unloading condition inhibits the nuclear importation of β-catenin (Case et al., 2008; Robinson et al., 2006), the current study then investigated whether the β-catenin function and its regulation are linked with MACF1 function under the normal and mechanical unloading conditions. Firstly, under the normal loading condition, β-catenin RNA and activated (non-phosphorylated) protein expression levels were both increased in MACF1-overexpression MC3T3-E1 osteoblastic cells when compared to negative control cells (Fig. 6A); and conversely, RNA and activated protein expression levels of β-catenin in MACF1-knockdown MC3T3-E1 cells were decreased when compared to the control cells (Fig. 6B). Secondly, the relationship between MACF1 on activation of β-catenin was further analyzed by using the TOPflash luciferase reporter assay which utilizes the TOPflash luciferase reporter plasmid having a LEF/TCF binding site to detect the activity of β-catenin. For control cells transfected by plasmid PEGFP-C1A, the relative luciferase activities of MACF1-knockdown MC3T3-E1 osteoblastic cells were significantly decreased compared to their negative control cells ( P<0.001), which suggested that the β-catenin activity in MACF1-knockdown osteoblastic cells was decreased. However, transfection of PEGFP-C1A-ACF7 plasmid in MACF1- knockdown cells significantly enhanced the β-catenin activity (P<0.01), when compared to the control cells (Fig. 6C). 3.6 Mechanical unloading condition down -regulated β-catenin through MACF1 To further investigate the relationship between MACF1 and β -catenin signaling under the mechanical unloading, the following studies were performed. Firstly, the effects of mechanical unloading (RPM treatment) on RNA and protein expression levels of β-catenin were examined in MC3T3-E1 cells. Results showed that after 24 and 48 hours of RPM treatment, β-catenin RNA level was down-regulated by 33% ( P<0.01) and 37.5% (P<0.001), respectively. Activated β-catenin protein levels were also decreased (Fig. 7A). Then, to determine β-catenin activities under mechanical unloading condition, the TOPflash luciferase reporter assay was carried out on RPM-treated or untreated MC3T3-E1 cells. The relative luciferase activities were significantly decreased in RPM treated cells compared with control cells (P<0.05, Fig. 7B), which suggested that β-catenin activities were decreased under mechanical unloading condition. Moreover, the expression level of β-catenin mRNA was found down-regulated by 43% in C57BL/6 mice femur tissue after 28 days HLU (P<0.001, Fig. 7C), which was consistent with the in vitro results and confirmed that mechanical unloading suppressed β-catenin signaling. Furthermore, when comparing MACF1-knockdown MC3T3-E1 cells and their related negative control cells, β-catenin RNA expression levels were found significantly decreased in negative control cells after RPM treatment ( P<0.001), but not in MACF1-knockdown MC3T3-E1 cells (Fig. 7D), suggesting that mech anical unloading may regulate β-catenin expression through MACF1. Finally, MACF1-overexpression (with PEGFP-C1A-ACF7 transfection) was found to significantly attenuate RPM treatment-induced reduction in β-catenin mRNA expression in MC3T3-E1 cells, when compared to the cells without MACF1-overexpression (transfected PEGFP-C1A) (P<0.01, Fig. 7E). 3.7 β- catenin attenuated the inhibitory effects of MACF1-knockdown or RPM on osteoblast proliferation Lastly, analyses were carried out to examine whether β-catenin agonist LiCl treatment can rescue the inhibitory effects of MACF1-knockdown or RPM on osteoblast proliferation. In MC3T3-E1 cells transfected with β-catenin siRNA, treatment of LiCl reversed the reduced both β-catenin activity and β-catenin activated protein level (Fig. 8A, D). RNA expression level of LEF1 and CyclinD1, downstream genes of β-catenin were significantly decreased by β-catenin siRNA, and recovered by LiCl treatment (Fig. 8B, C). Consistently, LiCl treatment significantly rescued the number of osteoblasts, which was significantly decreased by β-catenin siRNA treatment (P<0.05, Fig. 8G). The rescue effects of β -catenin on MACF1-knockdown MC3T3-E1 cells and RPM treated MC3T3-E1 were further analyzed. While both MACF1-knockdown and RPM reduced β-catenin expression in MC3T3-E1 cells, this effect was partially rescued by LiCl treatment (Fig. 8E, F). Consistently, while both MACF1-knockdown and RPM treatment reduced the numbers of MC3T3-E1 cells, LiCl treatment partially attenuated this effect (Fig. 8H, I). 4. Discussion MACF1 is a cytoskeletal protein that is ubiquitously expressed in different tissues (Hu et al., 2016; Suozzi et al., 2012) and correlated with various physiological and pathological effects (Hu et al., 2016; Suozzi et al., 2012). Previous studies have shown that MACF1 regulates actin and microtubule cytoskeletal dynamics (Wu et al., 2008; Wu et al., 2011; Yue et al., 2016) and has a role in regulating cardiomyocyte adaptation to hemodynamic overload (Fassett et al., 2013). However, the functions and associated mechanisms of MACF1 in response to mechanical unloading and in regulating proliferation of osteoblasts (responsible for bone formation) are unclear. In the present study, MACF1 expression was demonstrated to be decreased in osteoblasts under the mechanical unloading condition in both in vitro and in vivo mechanical unloading models. Mechanical unloading inhibited osteoblast proliferation, while MACF1 promoted osteoblast proliferation. Furthermore, by establishing the mechanical unloading condition on both MACF1-knockdown and overexpression MC3T3-E1 cells, we found that the inhibition of osteoblast proliferation under mechanical unloading was related to down-regulation of MACF1. These findings suggest that MACF1 may play an important role in the inhibitory effect of mechanical unloading on osteoblast proliferation. Moreover, we revealed that the facilitating effect of MACF1 to osteoblast proliferation may be through its effect in regulating β -catenin signaling activation. The current study has discovered that MACF1 may act as a sensor of mechanical unloading in osteoblasts and suggests that it responds to mechanical loading and positively regulate osteoblast proliferation via promoting β -catenin signaling. Osteoblasts respond to mechanical unloading (Hughes-Fulford & Lewis, 1996; Hughes-Fulford et al., 1998), and decreased osteoblast proliferation has been claimed to play an important part in the process of spaceflight mechanical unloading-induced bone loss (Morey & Baylink, 1978; Wronski & Morey, 1983). Our in vitro results proved that mechanical unloading conferred by RPM significantly inhibited osteoblast proliferation, which was consistent with previous reports. Sun et al. used clinorotat to simulate the mechanical unloading condition, and found that the proliferation rate of MC3T3-E1 osteoblastic cells was decreased by clinorotation treatment (Sun et al., 2015). Similarly, Hughes-Fulford et al. also found the proliferation of MC3T3-E1 cells decreased during space flight (Hughes-Fulford & Lewis, 1996), which again proved that mechanical unloading inhibits osteoblast proliferation. Mechanical unloading can affect osteoblast physiological behaviors through multiple molecular targets (Sankaran et al., 2016; Wang et al., 2013). MACF1, extensively expressed in osteoblasts, is a cytoskeletal protein regulating actin and microtubule cytoskeletal dynamics (Kodama et al., 2003; Wu et al., 2008; Wu et al., 2011; Yue et al., 2016), and it is sensitive to hemodynamic overload (Fassett et al., 2013) . Our previous study showed that MACF1 expression was altered under Large Gradient High Magnetic Field-induced mechanical unloading condition (Qian et al., 2009), which implied that MACF1 expression might correlate with mechanical unloading in osteoblasts. In our current study, RPM-induced mechanical unloading condition was found to decrease MACF1 expression in MC3T3-E1 cells, which was identical with our previous study. Furthermore, using a hind limb unloading (HLU) model which simulates in vivo mechanical unloading condition (Morey-Holton & Globus, 2002), the current study also observed the significantly decreased MACF1 expression levels in the femurs of C57BL/6 mice with HLU as well as in bone-derived MSCs, which are osteogenic cells isolated from cortical bone of HLU mice femur, implying that the mechanical unloading condition decreased MACF1 expression in osteoblasts in vivo. All these results suggest that MACF1 is sensitive to mechanical unloading stimulus. MACF1 regulates various cellular physiological processes (Leung et al., 1999; Leung et al., 2002). In Afghani’s study, MACF1 promoted proliferation of glioblastoma cells (Afghani et al., 2017). In the current study, we showed that MACF1 knockdown reduced proliferation of MC3T3-E1 osteoblastic cells, which was consistent with our previous work (Hu et al., 2015). Conversely, MACF1 overexpression was found to increase proliferation of the osteoblastic cells. Furthermore, restoration of MACF1 expression was found to rescue the reduction of proliferation in MACF1-knockdown cells. All of these findings suggest that MACF1 positively regulates osteoblast proliferation. Then, we investigated functions of MACF1 in mechanical unloading condition-induced reduction in osteoblast proliferation. By conducting RPM treatment on the established MACF1-knockdown MC3T3-E1 cells, we observed no significant effects of RPM on proliferation in cells with the deficiency of MACF1. This implied that MACF1 may act as a sensor to mechanical stimulation. Without its presence, the alteration of mechanical environment cannot be perceived by osteoblasts. Moreover, by conducting RPM experiment in cells with MACF1 overexpression, we also observed the rescue effects of MACF1 on mechanical unloading condition. The overexpression of MACF1 significantly enhanced proliferation of MC3T3-E1 cells even under the RPM treatment, which again demonstrated the role of MACF1 in sensing mechanical unloading and regulating osteoblast proliferation. However, the downstream molecular mechanisms for the MACF1 effect in regulating osteoblast proliferation remain unclear. Previously , MACF1 was shown to participate in Wnt/β-catenin pathway allowing β-catenin to dissociate from the β-catenin/AXIN/APC complex and thus transport to cell nucleus (Chen et al., 2006; Hu et al., 2017). Here, we found that the expression levels of β-catenin increased in MACF1-overexpression MC3T3-E1 cells, but decreased in MACF1-knockdown MC3T3-E1 cells. Furthermore, our TOPflash luciferase reporter assay also demonstrated a positive correlation between β -catenin activity and MACF1 expression, which is consistent with previous studies (Chen et al., 2006; Hu et al., 2017). These findings indicate that MACF1 positively regulates β-catenin signaling in osteoblasts. It has been reported that β -catenin was sensitive to mechanical stimuli, and mechanical unloading or overloading would affect the stability of β -catenin-AXIN complex and β-catenin nuclear importation (Case et al., 2008; Robinson et al., 2006). Furthermore, since the effects of mechanical stimuli on β-catenin regulation have been found to be correlated with the function of MACF1 (Chen et al., 2006), we proposed in the current study that the alteration of β-catenin under mechanical unloading circumstance may involve the function of MACF1 to perceive the mechanical environment. Here, while we observed that β -catenin expression and activation were significantly decreased under the mechanical unloading condition, the reduction was not observed in MACF1-knockdown osteoblastic cells. Conversely, MACF1-overexpression was found to significantly attenuate RPM treatment-induced reduction in β-catenin expression. These findings suggest that the effects of mechanical unloading to β-catenin significantly correlated with MACF1 expression in osteoblastic cells, and proved that MACF1 took an essential part in β-catenin regulation under mechanical unloading condition. Wnt/β-catenin pathway is a critical pathway to osteoblast physiology (Kang & Robling, 2014), and reports have shown that β-catenin regulates cell proliferation through activating LEF1 (Galindo et al., 2007; Wang et al., 2017). As an important osteogenic regulator, β-catenin has been reported to be phosphorylated by GSK-3β and then degraded by ubiquitination (Amit et al., 2002; Ikeda et al., 1998). Lithium, however, can inhibit the activities of GSK-3β thus enhance β-catenin activity (Silva et al., 2010). In our study, we also investigated the functions of β-catenin in osteoblast proliferation and examined the rescue effects of lithium chloride under the mechanical unloading condition. We observed that in β -catenin deficient MC3T3-E1 cells (as induced by β-catenin siRNA), lithium chloride treatment enhanced β-catenin function and further enhanced osteoblast proliferation. Similarly, lithium chloride rescue effects were also observed in MACF1-knockdown and RPM-treated cells. Thus, we can conclude from these observations that MACF1 responds to mechanical unloading and regulates osteoblast proliferation via β-catenin signaling. Taken together, a novel cell physiological mechanism for the relationship for mechanical unloading, MACF1 expression, β -catenin expression/activity, and osteoblast proliferation can be concluded from the results in the current study (Fig. 9). Mechanical unloading suppresses MACF1 expression in osteoblasts. MACF1, acting upstream of β -catenin, then regulates β-catenin expression and activity. Consequently, β -catenin responds to mechanical unloading stimuli through MACF1 and promotes osteoblast proliferation. In conclusion, the current study has revealed that MACF1, as a mechanical unloading sensitive cytosk eletal protein, promotes osteoblast proliferation via β-catenin. This study has uncovered a novel mechanical unloading sensitive factor, MACF1. Moreover, our study has discovered a new mechanism how mechanical unloading regulates osteoblast proliferation, and has provided new insights and potential targets for preventing and treating bone disorders caused by insufficient mechanical stimuli. References Afghani N, Mehta T, Wang J, Tang N, Skalli O, and Quick QA. 2017. Microtubule actin cross-linking factor 1, a novel target in glioblastoma. International Journal of Oncology 50(1), 310-316. Alzahrani MM, Anam EA, Makhdom AM, Villemure I, and Hamdy RC. 2014. The effect of altering the mechanical loading environment on the expression of bone regenerating molecules in cases of distraction osteogenesis. Frontiers in endocrinology 5 Amit S, Hatzubai A, Birman Y, Andersen JS, Ben-Shushan E, Mann M, Ben-Neriah Y, and Alkalay I. 2002. Axin-mediated CKI phosphorylation of β-catenin at Ser 45: a molecular switch for the Wnt pathway. Genes & development 16(9), 1066-1076. Burgo A, Proux-Gillardeaux V, Sotirakis E, Bun P, Casano A, Verraes A, Liem RK, Formstecher E, Coppey-Moisan M, and Galli T. 2012. A molecular network for the transport of the TI-VAMP/VAMP7 vesicles from cell center to periphery. Developmental cell 23(1), 166-180. Case N, Ma M, Sen B, Xie Z, Gross TS, and Rubin J. 2008. β -Catenin levels influence rapid mechanical responses in osteoblasts. Journal of Biological Chemistry 283(43), 29196-29205. Chen H, Lin C, Lin C, Perez-Olle R, Leung CL, and Liem RK. 2006. The role of microtubule actin cross-linking factor 1 (MACF1) in the Wnt signaling pathway. Genes & development 20(14), 1933-1945. Chen Z, Zhao F, Qi Y, Hu L, Li D, Yin C, Su P, Zhang Y, Ma J, and Qian J. 2016. Simulated microgravity alters the expression of cytoskeleton- and ATP-binding-related genes in MLO-Y4 osteocytes. Acta Astronautica 129186-192. Fassett JT, Xu X, Kwak D, Wang H, Liu X, Hu X, Bache RJ, and Chen Y. 2013. Microtubule Actin Cross-linking Factor 1 regulates cardiomyocyte microtubule distribution and adaptation to hemodynamic overload. PloS one 8(9), e73887. Galindo M, Kahler RA, Teplyuk NM, Stein JL, Lian JB, Stein GS, Westendorf JJ, and Van Wijnen AJ. 2007. Cell cycle related modulations in Runx2 protein levels are independent of lymphocyte enhancer-binding factor 1 (Lef1) in proliferating osteoblasts. Journal of molecular histology 38(5), 501-506. Gaudio A, Pennisi P, Bratengeier C, Torrisi V, Lindner B, Mangiafico RA, Pulvirenti I, Hawa G, Tringali G, and Fiore CE. 2010. Increased sclerostin serum levels associated with bone formation and resorption markers in patients with immobilization-induced bone loss. The Journal of Clinical Endocrinology & Metabolism 95(5), 2248-2253. Goryunov D, and Liem RK. 2016. Microtubule-Actin Cross-Linking Factor 1: Domains, Interaction Partners, and Tissue-Specific Functions. Methods in Enzymology 569331-353. Guo T, Wang W, Zhang H, Liu Y, Chen P, Ma K, and Zhou C. 2011. ISL1 promotes pancreatic islet cell proliferation. PloS one 6(8), e22387. Hu L, Qian A, Wang Y, Di S, and Shang P. 2013. Inhibitory effect of simulated microgravity on differentiating preosteoblasts. Advances in Space Research 51(1), 107-114. Hu L, Su P, Li R, Yan K, Chen Z, Shang P, and Qian A. 2015. Knockdown of microtubule actin crosslinking factor 1 inhibits cell proliferation in MC3T3-E1 osteoblastic cells. BMB Rep 48(10), 583-588. Hu L, Su P, Li R, Yin C, Zhang Y, Shang P, Yang T, and Qian A. 2016. Isoforms, structures, and functions of versatile spectraplakin MACF1. BMB reports 49(1), 37. Hu L, Su P, Yin C, Zhang Y, Li R, Yan K, Chen Z, Li D, Zhang G, and Wang L. 2017. Microtubule actin crosslinking factor 1 promotes osteoblast differentiation by promoting β-catenin/TCF1/Runx2 signaling axis. Journal of Cellular Physiology Hughes-Fulford M, and Lewis ML. 1996. Effects of microgravity on osteoblast growth activation. Experimental cell research 224(1), 103-109. Hughes-Fulford M, Tjandrawinata R, Fitzgerald J, Gasuad K, and Gilbertson V. 1998. Effects of microgravity on osteoblast growth. Gravitational and space biology bulletin: publication of the American Society for Gravitational and Space Biology 11(2), 51-60. Ikeda S, Kishida S, Yamamoto H, Murai H, Koyama S, and Kikuchi A. 1998. Axin, a negative regulator of the Wnt signaling pathway, forms a complex with GSK-3β and β-catenin and promotes GSK-3β-dependent phosphorylation of β-catenin. The EMBO journal 17(5), 1371-1384. Imbert L, Aurégan JC, Pernelle K, and Hoc T. 2014. Mechanical and mineral properties of osteogenesis imperfecta human bones at the tissue level. Bone 6518-24. Ingber DE. 1997. Tensegrity: the architectural basis of cellular EAPB02303 mechanotransduction. Annual Review of Physiology 59(1), 575-599.
Jiao K, Niu LN, Li QH, Ren GT, Zhao CM, Liu YD, Tay FCM, and Wang MQ. 2015. β2-adrenergic signal transduction plays a detrimental role in subchondral bone loss of temporomandibular joint in osteoarthritis. Scientific Reports 5(1), 12593.
Ka M, Jung EM, Mueller U, and Kim WY. 2014. MACF1 regulates the migration of pyramidal neurons via microtubule dynamics and GSK-3 signaling. Developmental biology 395(1), 4-18.
Kang KS, and Robling AG. 2014. New insights into Wnt–Lrp5/6–β-catenin signaling in mechanotransduction. Frontiers in endocrinology 5
Karakesisoglou I, Yang Y, and Fuchs E. 2000. An epidermal plakin that integrates actin and microtubule networks at cellular junctions. The Journal of cell biology 149(1), 195-208.
Kodama A, Karakesisoglou I, Wong E, Vaezi A, and Fuchs E. 2003. ACF7: an essential integrator of microtubule dynamics. Cell 115(3), 343-354.
Leung CL, Sun D, Zheng M, Knowles DR, and Liem RK. 1999. Microtubule actin cross-linking factor (MACF) a hybrid of dystonin and dystrophin that can interact with the actin and microtubule cytoskeletons. The Journal of cell biology 147(6), 1275-1286.
Leung CL, Green KJ, and Liem RK. 2002. Plakins: a family of versatile cytolinker proteins. Trends in cell biology 12(1), 37-45.
Liang Y, Shi C, Yang J, Chen H, Xia Y, Zhang P, Wang F, Han H, and Qin H. 2013. ACF7 regulates colonic permeability. International journal of molecular medicine 31(4), 861-866.
Louis KS, and Siegel AC. 2011. Cell viability analysis using trypan blue: manual and automated methods. Mammalian cell viability: methods and protocols 7-12.
Luo M, Yang Z, Li J, Xu H, Li S, Zhang W, Qian A, and Shang P. 2013. Calcium influx through stretch-activated channels mediates microfilament reorganization in osteoblasts under simulated weightlessness. Advances in Space Research 51(11), 2058-2068.
Morey-Holton ER, and Globus RK. 2002. Hindlimb unloading rodent model: technical aspects. Journal of applied physiology 92(4), 1367-1377.
Morey ER, and Baylink DJ. 1978. Inhibition of bone formation during space flight. Science 201(4361), 1138-1141.
Mundra JJ, Terskiy A, and Howells RD. 2012. Naltrindole inhibits human multiple myeloma cell proliferation in vitro and in a murine xenograft model in vivo. Journal of Pharmacology and Experimental Therapeutics 342(2), 273-287.
Oppl B, Michitsch G, Misof B, Kudlacek S, Donis J, Klaushofer K, Zwerina J, and Zwettler E. 2014. Low bone mineral density and fragility fractures in permanent vegetative state patients. Journal of Bone and Mineral Research 29(5), 1096-1100.
Papachroni KK, Karatzas DN, Papavassiliou KA, Basdra EK, and Papavassiliou AG. 2009. Mechanotransduction in osteoblast regulation and bone disease. Trends in molecular medicine 15(5), 208-216.
Pardo SJ, Patel MJ, Sykes MC, Platt MO, Boyd NL, Sorescu GP, Xu M, Van Loon JJWA, Wang MD, and Jo H. 2005. Simulated microgravity using the Random Positioning Machine inhibits differentiation and alters gene expression profiles of 2T3 preosteoblasts. American Journal of Physiology-cell Physiology 288(6),1211-1221.
Qian A, Hu L, Gao X, Zhang W, Di S, Tian Z, Yang P, Yin D, Weng Y, and Shang P. 2009. Large gradient high magnetic field affects the association of MACF1 with actin and microtubule cytoskeleton. Bioelectromagnetics 30(7), 545-555.
Qian AR, Li D, Han J, Gao XX, Di SM, Zhang W, Hu LF, and Shang P. 2012. Fractal Dimension as a Measure of Altered Actin Cytoskeleton in MC3T3-E1 Cells Under Simulated Microgravity Using 3-D/2-D Clinostats. IEEE Transactions on Biomedical Engineering 59(5), 1374-1380.
Robinson JA, Chatterjee-Kishore M, Yaworsky PJ, Cullen DM, Zhao W, Li C, Kharode Y, Sauter L, Babij P, and Brown EL. 2006. Wnt/β-catenin signaling is a normal physiological response to mechanical loading in bone. Journal of Biological Chemistry 281(42), 31720-31728.
Sankaran JS, Li B, Donahue LR, and Judex S. 2016. Modulation of unloading-induced bone loss in mice with altered ERK signaling. Mammalian Genome 27(1-2), 47-61.
Silva AK, Yi H, Hayes SH, Seigel GM, and Hackam AS. 2010. Lithium chloride regulates the proliferation of stem-like cells in retinoblastoma cell lines: a potential role for the canonical Wnt signaling pathway.
Sun Z, Cao X, Hu Z, Zhang L, Wang H, Zhou H, Li D, Zhang S, and Xie M. 2015. MiR-103 inhibits osteoblast proliferation mainly through suppressing Cav1. 2 expression in simulated microgravity. Bone 76121-128.
Suozzi KC, Wu X, and Fuchs E. 2012. Spectraplakins: master orchestrators of cytoskeletal dynamics. J Cell Biol 197(4), 465-475.
Van Loon JJ. 2007. Some history and use of the random positioning machine, RPM, in gravity related research. Advances in Space research 39(7), 1161-1165.
Wang F, Huang W, Hu X, Chen C, Li X, Qiu J, Liang Z, Zhang J, Li L, and Wang X. 2017. Transcription factor AP-2β suppresses cervical cancer cell proliferation by promoting the degradation of its interaction partner β-catenin. Molecular Carcinogenesis
Wang X, Guo B, Li Q, Peng J, Yang Z, Wang A, Li D, Hou Z, Lv K, and Kan G. 2013. miR-214 targets ATF4 to inhibit bone formation. Nature medicine 19(1), 93-100.
Wronski T, and Morey E. 1983. Effect of spaceflight on periosteal bone formation in rats. American journal of physiology-regulatory, integrative and comparative physiology 244(3), R305-R309.
Wu X, Kodama A, and Fuchs E. 2008. ACF7 regulates cytoskeletal-focal adhesion dynamics and migration and has ATPase activity. Cell 135(1), 137-148.
Wu X, Shen Q, Oristian DS, Lu CP, Zheng Q, Wang H, and Fuchs E. 2011. Skin stem cells orchestrate directional migration by regulating microtubule-ACF7 connections through GSK3β. Cell 144(3), 341-352.
Yue J, Zhang Y, Liang WG, Gou X, Lee P, Liu H, Lyu W, Tang W-J, Chen S-Y, and Yang F. 2016. In vivo epidermal migration requires focal adhesion targeting of ACF7. Nature communications 7
Zhu H, Guo Z, Jiang X, Li H, Wang X, Yao H, Zhang Y, and Mao N. 2010. A protocol for isolation and culture of mesenchymal stem cells from mouse compact bone. Nature protocols 5(3), 550-560.