Downregulated fat mass and obesity-associated protein inhibits bone resorption and osteoclastogenesis by nuclear factor-kappa B inactivation

Jinpeng Zhuang, Hua Ning, Maoqing Wang, Wei Zhao, Yongbin Jing, Xiaoqi Liu, Jianing Zu, Pengyu Kong, Xiaoyan Wang, Changhao Sun, Jinglong Yan
a Department of Orthopedic Surgery, The 2nd Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang Province 150001, People’s Republic of China
b Department of Nutrition and Food Hygiene, Public Health College, Harbin Medical University, Harbin, Heilongjiang Province 150081, People’s Republic of China

During osteoporosis, fat mass and obesity-associated protein (FTO) promotes the shift of bone marrow mesen- chymal stem cells to adipocytes and represses osteoblast activity. However, the role and mechanisms of FTO on osteoclast formation and bone resorption remain unknown. In this study, we investigated the effect of FTO on RAW264.7 cells and bone marrow monocytes (BMMs)-derived osteoclasts in vitro and observed the influence of FTO on ovariectomized (OVX) mice model to mimic postmenopausal osteoporosis in vivo. Results found that FTO was up-regulated in BMMs from OVX mice. Double immunofluorescence assay showed co-localization of FTO with tartrate-resistant acid phosphatase (TRAP) in femurs of OVX mice. FTO overexpression enhanced TRAP- positive osteoclasts and F-actin ring formation in RAW264.7 cells upon RANKL stimulation. The expression of osteoclast differentiation-related genes, including nuclear factor of activated T cells c1 (NFATc1) and c-FOS, wasupregulated in BMMs and RAW264.7 cells after FTO overexpression. FTO overexpression induced the phos- phorylation and nuclear translocation of factor-kappa B (NF-κB) p65 in BMMs and RAW264.7 cells exposed to RANKL. ChIP and dual-luciferase assays revealed that FTO overexpression contributed to RANKL-induced binding of NF-κB to NFATc1 promoter. Rescue experiments suggested that FTO overexpression-mediated oste-oclast differentiation was suppressed after intervention with a NF-κB inhibitor pyrrolidine dithiocarbamate.
Further in vivo evidence revealed that FTO knockdown increased bone trabecula and bone mineral density, inhibited bone resorption and osteoclastogenesis in osteoporotic mice. Collectively, our research demonstratesthat downregulated FTO inhibits bone resorption and osteoclastogenesis through NF-κB inactivation, whichprovides a novel reference for osteoporosis treatment.

1. Introduction
Osteoporosis is a systemic skeletal disease characterized by a loss in bone mass and bone mineral density (BMD), which is recognized as a global public health problem and a heavy socioeconomic burden [1]. This disease results in fragility fractures at typical sites, such as lumbar spine, femoral neck, and distal radius. The major pathogenesis of oste- oporosis is caused by a dysfunction in bone remodeling, which is due to excessive bone resorption and poor bone formation [2]. Postmenopausal women are prone to osteoporosis because estrogen deficiency results in increased bone resorption and decreased bone formation [3]. The cur- rent therapeutics of postmenopausal osteoporosis suppresses bone resorption by activation of bone remodeling and thereby prevents fractures. However, the development of a novel treatment of osteopo- rosis is still needed due to the limitations of anti-osteoporosis drugs.
The majority of studies indicate that obesity has a positive effect on BMD and skeletal strength [4]. Osteoblasts and adipocytes derive from bone marrow mesenchymal stem cells (BMSCs), which can trans- differentiate into each other. With aging, the BMSCs shift to adipocytes accompanied by osteoblast function decreases, and osteoclast activityincreases, which eventually arouse osteoporosis [5]. Previous evidence showed that fat mass and obesity-associated protein (FTO) affected not only obesity phenotypes, but also osteoporosis phenotypes, like BMD [6]. FTO is an RNA demethylase, which can demethylate a variety of methylated nucleic acids and play an important role in many biological and pathological processes, such as differentiation [7], obesity [8], and diabetes [9]. Sachse et al. [10] found that FTO demethylase activity was necessary for bone growth and bone mineralization of normal mice. The balance between adipocytes and osteoblasts differentiated from BMSCs was modulated by FTO. During the osteoporosis, FTO expression in bone marrow was up-regulated, which promoted the shift of BMSCs to adi- pocytes and repressed osteoblast activity [11]. However, the roles of FTO on osteoclasts in osteoporosis remain unknown.
The osteoclast is a multinucleated giant cell, which is derived from myeloid precursors belonging to the monocyte/macrophage family. Myeloid precursors differentiated into osteoclasts after stimulation by macrophage colony-stimulating factor 1 (M-CSF). Receptor activator ofnuclear factor κ-B ligand (RANKL) is a key extracellular regulator ofosteoclast differentiation and activation who can bind to the surfacereceptor RANK of osteoclast precursors, thereby activating nuclear fac- tor-κB (NF-κB), c-FOS, phospholipase Cγ (PLCγ), and nuclear factor of activated T cells c1 (NFATc1) to induce differentiation of precursors into osteoclasts [12]. A study showed that the loss of NF-κB suppressed TGF-β-induced osteoclast survival [13]. Another research suggested that FTO promoted NF-κB activation by positively regulating reactive oXygenspecies (ROS) production in pancreatic beta cells [14]. Therefore, wehypothesize that FTO might promote osteoclast formation by regulating the NF-κB pathway. To prove our conjecture, we established a mouse model of osteoporosis by ovariectomy to investigate the effects of FTOon osteoclast in osteoporotic mice, and further explore its underlying mechanism.

2. Materials and methods
2.1. Animal models
Eight-week-old female C57BL/6 J mice were purchased from Liaoning Changsheng biotechnology co., Ltd. (Benxi, China). All mice were housed under experimental environment: temperature of 25 ◦C,the humidity of 45–55%, light/dark cycle of 12/12 h and free access towater and food. This study was performed with the approval of the Ethics Committee of Harbin Medical University. All animal experiments were carried out according to the Guide for the Care and Use of Labo- ratory Animals. All mice were randomly divided into the OVX (mice receiving ovariectomy) group and the sham (mice not receiving ovari- ectomy) group (n 6/group). Mice were anesthetized with 50 mg/kg pentobarbital sodium. In the OVX group, an incision at one-third ofmice’ bilateral lower abdomen was made, and then bilateral ovarieswere completely removed. Subsequently, the incision was sutured with a hemostatic suture. In the sham group, the mesentery of the mice was removed, with weight equal to the ovaries, and then the same proced- ures were performed as the OVX group. After 12 weeks, the femurs were acquired from the sham and OVX mice.

2.2. Bone marrow-derived macrophage (BMM) isolation and osteoclast differentiation
BMMs were isolated from the femurs of mice as previously described [15]. Briefly, the end of each femur was cut off, and the bone marrow cells were collected by flushing the bone marrow cavity. Bone marrow cells were incubated in the complete alpha minimum essential medium (α-MEM, Procell, Wuhan, China) containing 10% fetal bovine serum(FBS) overnight. Non-adherent cells were harvested and then incubated in α-MEM supplemented with 10% FBS and 50 ng/mL M-CSF for 3 days to acquire adherent BMMs. To obtain BMM-derived osteoclasts, BMMs were cultured in α-MEM containing 10% FBS, 20 ng/mL M-CSF and 50 ng/mL RANKL for 7 days with or without pyrrolidine dithiocarbamate (PDTC; 100 μM; MCE, Shanghai, China) 24 h after lentivirus infection.

2.3. RAW264.7 cell culture and osteoclast differentiation
RAW264.7 cells were purchased from Procell and cultured in Dul- becco’s Modified Eagle Medium (DMEM; Gibco, California, USA) con- taining 10% FBS at 37 ◦C with 5% CO2. When reached 90% confluency,the cells were treated with trypsin (0.2%, Zhongqiao Xinzhou, Shanghai, China) and incubated with DMEM solution for the termination of digestion. Next, the cells were centrifuged at 90 g for 3 min at 4 ◦C andthe pellet was resuspended in DMEM, after which incubated in a 37 ◦Cincubator with 5% CO2 for 24 h. Subsequently, RAW264.7 cells were treated with RANKL (100 ng/mL, Sino Biological, Beijing, China) toinduce osteoclast differentiation with or without PDTC (100 μM) in thefresh medium 24 h after lentivirus infection. Five days later, RAW264.7 cell-derived osteoclasts were harvested.

2.4. In vitro lentiviral infection
A lentiviral vector encoding short hairpin RNA targeting FTO [len- tiviral vector shRNA-FTO (shFTO-LV)] or FTO-overexpressed lentiviral vector (over-FTO-LV), and its negative control lentiviral vector (NC-LV/ vector-LV) were constructed and then infected BMMs and RAW264.7cells for 24 h, followed by incubating at 37 ◦C with 5% CO2 for 48 h.
Then, cells in the culture media were collected.

2.5. In vivo delivery of lentivirus
Osteoporotic mice were randomly divided into two groups (n 6/ group): a shFTO-LV group (mice in the OVX group infected with shFTO- LV), and an NC-LV group (mice in the OVX group infected with NC-LV). Four weeks after osteoporosis, the femoral medullary cavity of the rightside of mice in the shFTO-LV group was periosteally injected with the shFTO lentivirus (1 107 TU/10 μL) once every 4 weeks. After 12 weeks, mice in each group were euthanasia and femur tissues werecollected.

2.6. Immunohistochemistry
The femur tissues were fiXed in 4% (w/v) paraformaldehyde (PFA), embedded in paraffin, sectioned in 5 μm thick, deparaffinized in xylene and rehydrated in graded alcohols. After treating with 3% hydrogenperoXide for 15 min, the sections were blocked with goat serum (Solarbio, Beijing, China) for 15 min at room temperature, and then incubated using primary antibodies, including rabbit anti-RANKL (dilution in 1: 50 PBS; Abclonal, Wuhan, China) and rabbit anti- osteoprotegrin (OPG; dilution in 1: 50 PBS; Abclonal), overnight at4 ◦C. Subsequently, the sections were cultured with horseradish peroX-idase (HRP)-conjugated goat anti-rabbit IgG secondary antibody (dilu- tion in 1: 500 PBS; Thermo Fisher, Waltham, MA, USA) for 30 min at room temperature. The sections were then stained with dia- minobenzidine, and counterstained with hematoXylin. After mounting with neutral balsam, target proteins were visualized under light mi-croscopy at original magnification ×400.

2.7. Immunofluorescence
BMMs and RAW264.7 cells were fiXed in 4% PFA, permeabilized with 0.1% triton X-100 and blocked with goat serum for 15 min at room temperature. Afterwards, femur sections, BMMs and RAW264.7 cells were incubated using primary antibodies, including rabbit anti-tartrate- resistant acid phosphatase (TRAP; dilution in 1: 50 PBS; Proteintech, Beijing, China), mouse anti-FTO (dilution in 1: 50 PBS; Santa, Arizona, USA) and rabbit anti-NF-kB p65 (dilution in 1: 200 PBS; Affinity,Changzhou, China) overnight at 4 ◦C. After washing with PBS, thesamples were incubated for 60–90 min with Cy3-conjugated goat anti- rabbit IgG (dilution in 1:200 PBS; Beyotime, Shanghai, China) or FITC-conjugated goat anti-mouse IgG (dilution in 1:200 PBS; Beyotime)antibody. Subsequently, bound primary antibodies were counterstained with 4′, 6-diamidino-2-phenylindole (DAPI; Aladdin, Shanghai, China) for cell count. After mounting with anti-fluorescence quenching agent,the immunofluorescence was observed by fluorescence microscopy (Olympus, Tokyo, Japan) at original magnification 400. TRAP- fluorescence intensity was quantified using Image-pro plus 6.0 (Media cybernetics, Rockville, MD, USA).

2.8. Chromatin immunoprecipitation (ChIP) assay
ChIP assay was performed using RAW264.7 cells treated with RANKL

2.9. Luciferase assay
RAW264.7 cells were transfected with 0.4 μg of pGL3-basic plasmid containing mouse NFATc1 promoter regions, mouse p65 overexpressionplasmid and pRL-TK plasmid using Attractene Transfection Reagent (QIAGEN, Duesseldorf, Germany). After 4 h incubation, cells were recultured in the fresh medium for 24 h, after which incubated with RANKL (100 ng/mL). Relative luciferase activity was detected using a luciferase assay kit (KeyGEN, Nanjing, China) and normalized to Renilla luciferase activity.

2.10. Western blot
Total protein of the BMMs and RAW264.7 cells was extracted usingthe RIPA lysate (Beyotime). Protein concentrations were measured by the BCA Protein Assay Kit (Beyotime). The protein samples (20 μg) were separated on SDS-PAGE (Beyotime) and transferred onto polyvinylidenefluoride (PVDF, Thermo Fisher) membrane. Then, the membranes were blocked with 5% (m/v) bull serum albumin (BSA; Biosharp, Hefei, China) for 1 h and incubated overnight at 4 ◦C via using the followingdiluted primary antibodies: rabbit anti-FTO (dilution in 1: 2000 BSA; Proteintech), rabbit anti-NF-κB p65 (dilution in 1: 2000 BSA; Pro- teintech), rabbit anti-p-NF-κB p65 (ser276) (dilution in 1: 1000 BSA;Affinity), rabbit anti-IκBα (dilution in 1: 1000 BSA; CST, Boston, USA), rabbit anti-p-IκBα (ser32/ser36) (dilution in 1: 1000 BSA; Affinity),rabbit anti-NFATc1 (dilution in 1: 1000 BSA; ABclonal), rabbit anti-c- FOS (dilution in 1: 1000 BSA; ABclonal), and rabbit anti-β-actin (dilu- tion in 1: 2000 BSA; Proteintech). After washing with Tris-buffered sa-line Tween-20 (TBST) four times (5 min each), membranes were incubated at 37 ◦C for 40 min with HRP-labeled goat anti-rabbit andgoat anti-mouse (dilution in 1: 10000 BSA; Proteintech) IgG antibodies. Then, membranes were colorized by electrochemiluminescence (7 Sea biotech, Shanghai, China) and analyzed via the Gel-Pro-Analyzer 4.0 (Media Cybernetics).

2.11. Hematoxylin and eosin (HE) staining
The femurs were fiXed in 4% (w/v) PFA for 24 h and embedded in paraffin. Then, HE staining was conducted according to the previous study [16]. Histological sections of the femur of the mice in each group were observed by light microscopy (Olympus) at original magnification×40.
TritonX-100 for 30 min. Next, tissues and cells were incubated in the TRITC Phalloidin (Solarbio). Cell nucleus was stained using DAPI. After mounting with anti-fluorescence quenching agent, the fluorescence signal was observed by fluorescence microscopy (Olympus) at original magnification 400. The number of F-actin rings was counted in three visual fields randomly.
TRAP staining was performed using Acid Phosphatase, Leukocyte (TRAP) Kit (Sigma-Aldrich, St. Louis., MO, USA). Briefly, TRAP solution was added to the well and cultured with the cells at 37 ◦C for 1 h. Images of TRAP staining were observed under light microscopy (Olympus). The number of TRAP-positive cells was counted in a blinded manner at least three times. The parameters measured for bone resorption were furthercalculated: including osteoclast number per bone perimeter (OC⋅N/ B⋅Pm), and osteoclast number per bone surface (OC⋅N/BS).

2.12. Skeletal phenotyping
The distal ends of intact femurs from mice were scanned using micro- computed tomography (micro-CT, quantum GX, PerkinElmer) to assess bone mass, density, and trabecular microarchitecture. The following parameters obtained for the bone formation were calculated from these data: including BMD, bone volume against tissue volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), and trabecular separation (Tb.Sp.).

2.13. F-actin ring staining and TRAP staining
The localization of F-actin was used to detect osteoclast differentia- tion. FiXed tissues and RAW264.7 cells were cultured with 0.1%(100 ng/mL) with or without over-FTO-LV and shFTO-LV, according to the manufacturer’s instructions of ChIP Assay Kit (Wanleibio, Shenyang,China). Target sequences of NFATc1 promoter sequences were identified by PCR (40 cycles at 95 ◦C for 20 s, 55 ◦C for 20 s, 72 ◦C for 30 s) using primer pairs amplifying NFATc1-specific promoter regions with NF-κB-binding sequence, forward 5’-ACGCCCATGCAATCTGTTAG-3′ andreverse 5’-TTGCATGCTGAAGTCATTATGT-3′.

2.14. Statistical analysis
All data were shown as mean standard deviation (SD). Statistical analysis was performed using two-tailed unpaired Student’s t-test for two groups or one-way ANOVA followed by Tukey’s for multiple groups.
Data were carried out using GraphPad Prism software (Version 8.0). P <0.05 was considered statistically significant. 3. Results 3.1. Highly expressed FTO and increased osteoclastogenesis are observed in femurs of OVX mice Micro-CT analysis suggested that BMD was decreased in femurs of the OVX mice compared with the sham mice (Fig. 1A, P < 0.05). Next, we investigated the relative protein expression of FTO in BMMs withinthe femurs using western blot. Compared with the sham group, theprotein expression of FTO was significantly up-regulated in the OVX group (Fig. 1B, P < 0.05). Histological sagittal sections of the femurs revealed an obvious reduction in bone trabecula of the OVX mice rela-tive to the sham mice, and the bony meshwork appeared in femurs of the OVX mice (Fig. 1C). Since OPG and RANKL are crucial regulators of osteoclastogenesis [17], the expression of RANKL and OPG in the sham mice and the OVX mice was assessed using immunohistochemical assay. As shown in Fig. 1D, increased RANKL expression and decreased OPG expression were observed in femurs of the OVX mice compared with the sham mice. TRAP staining was used to determine the differentiatedmultinucleated osteoclasts. TRAP-positive osteoclasts were significantly up-regulated in femurs of the OVX mice (Fig. 1E, P < 0.05). Immuno- fluorescence analysis confirmed increase levels of FTO and TRAP in theOVX group. Double labeling for TRAP/FTO revealed an intense coloc- alization (Fig. 1F). 3.2. FTO affects TRAP expression in BMMs upon exposure to M-CSF and RANKL We constructed over-FTO-LV or shFTO-LV vector to infect BMMs,after which acquired mature osteoclasts upon exposure to M-CSF (20 ng/mL) and RANKL (50 ng/mL). As described in Fig. 2A, the expression of FTO protein level was significantly up-regulated in BMMs infected with over-FTO-LV, while down-regulated in BMMs infected with shFTO-LV (P < 0.05). Immunofluorescence analysis was further performed toassess the effect of FTO on TRAP expression. Results of immunofluo- rescence revealed that overexpression of FTO efficiently enhanced TRAP-fluorescence intensity in BMMs upon M-CSF and RANKL stimu-lation, while it was inhibited by FTO knockdown (Fig. 2B, P < 0.05). 3.3. FTO affects NF-κB and the downstream c-Fos/NFATc1 signaling pathways RANKL binding of RANK activates a series of signaling pathways, including NF-κB and the downstream c-FOS/NFATc1, which control theosteoclast formation and function [18]. To understand the influence ofFTO on osteoclast differentiation-related signaling pathways, we detected the protein expression of IκBα, NF-κB p65, NFATc1 and c-FOS in BMMs upon RANKL and M-CSF stimulation. Results of western blot showed that the protein levels of p-IκBα and p-NF-κB p65 were signifi- cantly increased but the level of IκBα was decreased in BMMs infected with over-FTO-LV upon RANKL and M-CSF stimulation (Fig. 3A, P < 0.05). On the contrary, infection with shFTO-LV significantly inhibited the phosphorylation levels of IκBα and NF-κB p65 (Fig. 3A, P < 0.05). Results of p65-immunofluorescence staining suggested that FTO over- expression promoted the translocation of NF-κB p65 to nuclear, whereasFTO knockdown repressed the nuclear translocation (Fig. 3B). In addi- tion, the expression of NFATc1 and c-FOS was remarkably up-regulated in BMMs infected with over-FTO-LV upon RANKL and M-CSF stimula-tion, while infection with shFTO-LV down-regulated the NFATc1 and c- FOS expression (Fig. 3C, P < 0.05). Rescue experiment was performed to uncover whether the NF-κB signaling regulated TRAP expression. As shown in Fig. 3D, overexpression of FTO evidently enhanced TRAP-fluorescence intensity, which was curbed by NF-κB inhibitor PDTC (P< 0.05). These findings implied that FTO might contribute to osteoclastformation. 3.4. The effect of FTO on differentiation of Raw264.7 cells to osteoclasts To figure out whether FTO induced osteoclast differentiation, pre- osteoclastic RAW264.7 cells were induced to differentiate into osteo- clasts by RANKL. As shown in Fig. 4A, protein expression of FTO was greatly up-regulated in RAW264.7 cells infected with over-FTO-LV, whereas FTO expression was down-regulated in RAW264.7 cells infec-ted with shFTO-LV (P < 0.05). Overexpression of FTO efficiently pro-moted differentiation of RAW264.7 cells to osteoclasts upon exposure to RANKL, exhibited by the increased number of TRAP-positive cells(Fig. 4B, P < 0.05). In contrast, knockdown of FTO significantly decreased TRAP-positive cells (Fig. 4B, P < 0.05). Furthermore, weobserved clearer F-actin ring structures in over-FTO-LV group compared with the vector-LV group, but knockdown of FTO markedly decreasedwas significantly increased in FTO-overexpressed RAW264.7 cells upon exposure to RANKL (Fig. 5A and B, P < 0.05). In contrast, silencing of FTO suppressed the phosphorylation of IκBα and NF-κB p65, as well asexpression of NFATc1 and c-FOS (Fig. 5A and B, P < 0.05). Results ofp65-immunofluorescence showed that FTO overexpression contributed to the nuclear translocation of NF-κB p65 in RAW264.7 cells exposed to RANKL, but this translocation was dampened by FTO knockdown(Fig. 5C). ChIP analysis revealed that FTO overexpression contributed to RANKL-induced binding of NF-κB to NFATc1 promoter, but this effect was reversed by knocking down FTO (Fig. 5D). Besides, results of dual-luciferase demonstrated that p65 increased RANKL-induced transcrip- tional activity of NFATc1 promoter (Fig. 5E, P < 0.05). To verify the role of FTO in the activation of NF-κB pathway, NF-κB inhibitor PDTC was applied to decrease the NF-κB level in FTO-overexpressed RAW264.7 cells. The results of TRAP staining and F-actin ring staining revealed thatPDTC alleviated osteoclast differentiation and decreased F-actin ring formation in FTO-overexpressed RAW264.7 cells, which indicated thatthe positive effect of FTO on osteoclast formation was regulated by the NF-κB activation (Figs. 5F, P < 0.05). 3.5. FTO induces Raw264.7 cell-derived osteoclastogenesis by activating NF-κB pathway To further explore the biological role and mechanism of FTO in os-teoclasts differentiated from RAW264.7 cells, we investigated whether FTO overexpression or silencing could affect the NF-κB/NFATc1 pathway involved in osteoclast differentiation. Results of western blot showed that the expression of p-IκBα, p-NF-κB p65, NFATc1 and c-FOSBMMs (Fig. 6A, P < 0.05). Images of HE staining exhibited that bony meshwork was decreased and bone trabecula was increased in osteo-porotic mice injected with shFTO-LV (Fig. 6B). Micro-CT analysis of femurs suggested that periosteal injection of shFTO-LV greatly increased the levels of BMD, BV/TV, Tb.N and Tb.Th compared to the NC-LV group, whereas the Tb.Sp level was down-regulated in shFTO-LVgroup (Figs. 6C-H, P < 0.05). Besides, quantification of the TRAPstaining showed a significantly decreased osteoclast number, perimeter and surface area in shFTO-LV group (Figs. 6I-K, P < 0.05). The result of F-actin ring staining exhibited that knockdown of FTO destroyed ringstructure and inhibited osteoclastogenesis in mice (Fig. 6L). 3.6. Knockdown of FTO inhibits bone resorption and osteoclastogenesis in femurs of osteoporotic mice Osteoporotic mouse model and shFTO-LV were used to further assessthe effect of FTO on osteoporosis in vivo. Western blot showed that knockdown of FTO significantly down-regulated FTO protein level inthe number of F-actin rings (Fig. 4C, P < 0.05). These data indicated that FTO induced the differentiation of Raw264.7 cells to osteoclasts. 4. Discussion Bone homeostasis is maintained by the balance between osteoclast- mediated bone resorption and osteoblast-mediated bone formation, while imbalance of this process can evoke osteoporosis. During osteo- porosis, FTO promotes the shift of BMSCs to adipocytes and suppresses osteoblast activity [11]. However, the role and mechanisms of FTO onosteoclast differentiation and bone resorption were not illuminated. In this study, we determine that FTO, a N6-methyladenosine (m6A) RNAdemethylase, is highly expressed in BMMs from femurs of osteoporotic mice. Moreover, FTO deficiency inhibits bone resorption and osteo-clastogenesis through NF-κB inactivation. A previous study has confirmed that FTO gene participates in the regulation of diverse physiological and pathological processes [8]. It has been reported that obesity-associated risk allele carriers of FTO gene exhibit dose-dependent increases in body mass index with aging [19]. A commonly carried allele of FTO is reported to be related to decreased brain volume in healthy elderly people [20]. These studies imply that FTO is one of the core mechanisms in age-related metabolic diseases, such as osteoporosis. A recent study has shown that FTO is up-regulated in the bone specimens of aged patients with osteoporosis [11]. In agreement with this study, we find that FTO is highly expressed in BMMs from femurs of osteoporotic mice. Besides, our study exhibits for the first time co-localization of FTO with TRAP-labeled osteoclasts in osteopo- rotic mice. Knocking down FTO decreased the fluorescence intensity of TRAP in BMMs upon exposure to M-CSF and RANKL, implying that FTO might contribute to osteoclast formation. To better understand the effect of FTO on osteoclast differentiation, we increase or silence theexpression of FTO in RAW264.7 cells upon exposure to RANKL. The results of TRAP staining suggest that FTO overexpression increases the number of TRAP-positive cells, whereas it is alleviated by FTO silencing. For osteoclast-mediated bone resorption, the formation of F-actin rings is considered as an important visual phenotype of mature osteoclasts [21]. The results of F-actin staining show an increased F-actin rings in FTO-overexpressed RAW264.7 cells, which confirm that FTOoverexpression stimulates osteoclast differentiation. Therefore, wedetermine that FTO overexpression contributes to osteoclast differentiation. The m6A is the most prevalent messenger RNA (mRNA) modification that has recently been proven to exert a crucial role in the regulation ofcell differentiation [22]. The dynamic m6A RNA modification is regu- lated by methyltransferases (m6A “writers”), demethylases (m6A“erasers”) and specific binding proteins (m6A “readers”) [22]. A previ- ous study suggested that knockdown of methyltransferase-like 3 (METTL3) inhibited osteoclast differentiation by enhancing the stabilityof Atp6v0d2 mRNA and blocking the NF-κB activation [23], indicating the important role of m6A modification in osteoclast differentiation. FTO has been found to stabilize MYC mRNA by inhibition of m6A reader YTHN6-methyladenosine RNA binding protein 2 (YTHDF2)-mediated degradation, thereby promoting leukemogenesis [24]. Furthermore, MYC increases osteoclast formation by inducing the expression ofNFATc1 [25]. In this study, we observe that FTO induces osteoclast differentiation, which may be due to decreased m6A abundance on the 5′-terminal and internal exons of MYC mRNA [24]. Interestingly, human and murine osteoclasts have been demon- strated to require NF-κB for survival, suggesting that targeting this survival factor is a frequent mechanism to maintain the survival of os-teoclasts [26]. A previous study has indicated that FTO overexpression activates NF-κB pathway through ROS generation in pancreatic β cells [14]. Besides, knockdown of FTO inhibited NF-κB signaling pathway and decreased mRNA stability of signal transducer and activator oftranscription 1 (STAT1) and peroXisome proliferation-activated recep- tor-γ (PPAR-γ) in macrophages through an YTHDF2-dependent pathway [27]. In this study, FTO-mediated phosphorylation of NF-κB p65 is observed in BMMs and RAW264.7 cells upon exposure to RANKL. Be- sides, the loss of total IκBα indicates the occurrence of IκBα phosphor- ylation in osteoclasts, since phosphorylation results in proteasome- mediated degradation of IκBα [28]. FTO overexpression remarkably induces the nuclear translocation of NF-κB p65. Thus, we confirm thatupon RANKL stimulation, the overexpression of FTO promotes the phosphorylation of NF-κB p65 and subsequent nuclear translocation, whereas the silencing of FTO attenuates this phenomenon. NF-κB modulates the expressions of osteoclastogenesis-related transcriptionfactors, such as c-Fos [29] and NFATc1 [30] in RANKL-mediated osteoclastogenesis. NFATc1 serves as a master regulator of osteoclasto- genesis by activating expression of osteoclast-specific genes, such as TRAP, CtsK, and DC-STAMP [31]. Guo et al. find that histone deacety- lase inhibitor CI-994 represses osteoclastogenesis through inhibiting NF-κB and the downstream c-Fos/NFATc1 signaling pathways [18]. Simi-larly, the reduction of c-Fos and NFATc1 expression was observed inBMMs and RAW264.7 after FTO silencing. To determine the role of NF- κB signaling in FTO-mediated promotion of osteoclast survival, we inspect response in osteoclasts in the presence of PDTC that can inhibit NF-κB. We find that the inhibition of NF-κB significantly blocks FTO- mediated osteoclast differentiation, as exhibited by decreased TRAP-positive cells and F-actin rings. These findings confirm that the in- crease of osteoclast differentiation caused by FTO overexpression ismediated by NF-κB activation. To verify the FTO-mediated promotion ofosteoclast differentiation, here, we examine the effect of FTO knock- down on the bone resorption and osteoclast differentiation in mice. We find that FTO knockdown inhibits the osteoclast differentiation of mouse BMMs. In addition, FTO deficiency increases BMD and inhibits bone resorption. EXcessive bone resorption results to the loss of bone mass and the defects in bone formation [32]. Therefore, FTO is considered as a crucial factor to maintain bone homeostasis. All these indicate the importance of FTO for bone resorption and osteoclastogenesis, which show that inhibition of FTO may be a potential therapeutic gene for osteoporosis. In summary, this study provides evidence that FTO-mediated pro- motion of osteoclast differentiation is regulated by NF-κB activation. Furthermore, knockdown of FTO by periosteal injection of lentiviralshRNA-FTO increases bone mass and BMD, inhibits bone resorption and osteoclastogenesis in osteoporotic mice. Our study provided a new method of osteoporosis treatment by downregulation of FTO expression using a lentiviral shRNA-FTO. References [1] V.K. Yadav, S. Balaji, P.S. Suresh, X.S. Liu, X. Lu, Z. Li, X.E. Guo, J.J. Mann, A.K. Balapure, M.D. Gershon, R. Medhamurthy, M. Vidal, G. Karsenty, P. Ducy,Pharmacological inhibition of gut-derived serotonin synthesis is a potential bone anabolic treatment for osteoporosis, Nat. Med. 16 (3) (2010) 308–312. [2] J.P. Bidwell, M.B. Alvarez, M. Hood Jr., P. Childress, Functional impairment of bone formation in the pathogenesis of osteoporosis: the bone marrow regenerativecompetence, Curr. Osteoporos. Rep. 11 (2) (2013) 117–125. [3] R. Nikander, H. Sievanen, A. Heinonen, R.M. Daly, K. Uusi-Rasi, P. Kannus, Targeted exercise against osteoporosis: a systematic review and meta-analysis for optimising bone strength throughout life, BMC Med. 8 (2010) 47. [4] J.P. Kemp, A. Sayers, G.D. Smith, J.H. Tobias, D.M. Evans, Using mendelian randomization to investigate a possible causal relationship between adiposity and increased bone mineral density at different skeletal sites in children, Int. J.Epidemiol. 45 (5) (2016) 1560–1572. [5] H. Jing, L. Liao, Y. An, X. Su, S. Liu, Y. Shuai, X. Zhang, Y. Jin, Suppression of EZH2 prevents the shift of osteoporotic MSC fate to adipocyte and enhances boneformation during osteoporosis, Mol. Ther. 24 (2) (2016) 217–229. [6] Y. Guo, H. Liu, T.L. Yang, S.M. Li, S.K. Li, Q. Tian, Y.J. Liu, H.W. Deng, The fat mass and obesity associated gene, FTO, is also associated with osteoporosis phenotypes, PLoS One 6 (11) (2011), e27312. [7] S. Geula, S. Moshitch-Moshkovitz, D. Dominissini, A.A. Mansour, N. Kol,M. Salmon-Divon, V. Hershkovitz, E. Peer, N. Mor, Y.S. Manor, M.S. Ben-Haim,E. Eyal, S. Yunger, Y. Pinto, D.A. Jaitin, S. Viukov, Y. Rais, V. Krupalnik,E. Chomsky, M. Zerbib, I. Maza, Y. Rechavi, R. Massarwa, S. Hanna, I. Amit, E.Y. Levanon, N. Amariglio, N. Stern-Ginossar, N. Novershtern, G. Rechavi, J.H. Hanna, Stem cells. m6A mRNA methylation facilitates resolution of naive pluripotency toward differentiation, Science 347 (6225) (2015) 1002–1006. [8] T.M. Frayling, N.J. Timpson, M.N. Weedon, E. Zeggini, R.M. Freathy, C.M. Lindgren, J.R. Perry, K.S. Elliott, H. Lango, N.W. Rayner, B. Shields, L.W. Harries, J.C. Barrett, S. Ellard, C.J. Groves, B. Knight, A.M. Patch, A.R. Ness,S. Ebrahim, D.A. Lawlor, S.M. Ring, Y. Ben-Shlomo, M.R. Jarvelin, U. Sovio, A.J. Bennett, D. Melzer, L. Ferrucci, R.J. Loos, I. Barroso, N.J. Wareham, F. Karpe, K.R. Owen, L.R. Cardon, M. Walker, G.A. Hitman, C.N. Palmer, A.S. Doney, A.D. Morris, G.D. Smith, A.T. Hattersley, M.I. McCarthy, A common variant in theFTO gene is associated with body mass index and predisposes to childhood and adult obesity, Science 316 (5826) (2007) 889–894. [9] F. Shen, W. Huang, J.T. Huang, J. Xiong, Y. Yang, K. Wu, G.F. Jia, J. Chen, Y.Q. Feng, B.F. Yuan, S.M. Liu, Decreased N(6)-methyladenosine in peripheral blood RNA from diabetic patients is associated with FTO expression rather than ALKBH5,J. Clin. Endocrinol. Metab. 100 (1) (2015) E148–E154. [10] G. Sachse, C. Church, M. Stewart, H. Cater, L. Teboul, R.D. CoX, F.M. Ashcroft, FTOdemethylase activity is essential for normal bone growth and bone mineralization in mice, Biochim. Biophys. Acta Mol. basis Dis. 1864 (3) (2018) 843–850. [11] G.S. Shen, H.B. Zhou, H. Zhang, B. Chen, Z.P. Liu, Y. Yuan, X.Z. Zhou, Y.J. Xu, The GDF11-FTO-PPARgamma axis controls the shift of osteoporotic MSC fate to adipocyte and inhibits bone formation during osteoporosis, Biochim. Biophys. ActaMol. basis Dis. 1864 (12) (2018) 3644–3654. [12] H. Takayanagi, Osteoimmunology: shared mechanisms and crosstalk between theimmune and bone systems, Nat. Rev. Immunol. 7 (4) (2007) 292–304. [13] A. Gingery, E.W. Bradley, L. Pederson, M. Ruan, N.J. Horwood, M.J. Oursler, TGF- beta coordinately activates TAK1/MEK/AKT/NFkB and SMAD pathways topromote osteoclast survival, EXp. Cell Res. 314 (15) (2008) 2725–2738. [14] H.Q. Fan, W. He, K.F. Xu, Z.X. Wang, X.Y. Xu, H. Chen, FTO inhibits insulin secretion and promotes NF-kappaB activation through positively regulating ROS production in pancreatic beta cells, PLoS One 10 (5) (2015), e0127705. [15] R. Xu, X. Shen, Y. Si, Y. Fu, W. Zhu, T. Xiao, Z. Fu, P. Zhang, J. Cheng, H. Jiang, MicroRNA-31a-5p from aging BMSCs links bone formation and resorption in the aged bone marrow microenvironment, Aging Cell 17 (4) (2018), e12794. [16] T.H. Huang, R.S. Yang, S.S. Hsieh, S.H. Liu, Effects of caffeine and exercise on the development of bone: a densitometric and histomorphometric study in youngwistar rats, Bone 30 (1) (2002) 293–299. [17] G. Silvestrini, P. Ballanti, F. Patacchioli, M. Leopizzi, N. Gualtieri, P. Monnazzi,E. Tremante, D. Sardella, E. Bonucci, Detection of osteoprotegerin (OPG) and itsligand (RANKL) mRNA and protein in femur and tibia of the rat, J. Mol. Histol. 36 (1–2) (2005) 59–67. [18] D. Guo, D. Hong, P. Wang, J. Wang, L. Chen, W. Zhao, L. Zhang, C. Yao, B. Chu,S. Chen, Z. Li, H. Chen, Histone deacetylase inhibitor CI-994 inhibits osteoclastogenesis via suppressing NF-kappaB and the downstream c-Fos/NFATc1signaling pathways, Eur. J. Pharmacol. 848 (2019) 96–104. [19] Y.F. Chuang, T. Tanaka, L.L. Beason-Held, Y. An, A. Terracciano, A.R. Sutin,M. Kraut, A.B. Singleton, S.M. Resnick, M. Thambisetty, FTO genotype and aging: pleiotropic longitudinal effects on adiposity, brain function, impulsivity and diet,Mol. Psychiatry 20 (1) (2015) 133–139. [20] A.J. Ho, J.L. Stein, X. Hua, S. Lee, D.P. Hibar, A.D. Leow, I.D. Dinov, A.W. Toga, A.J. Saykin, L. Shen, T. Foroud, N. Pankratz, M.J. Huentelman, D.W. Craig, J.D. Gerber, A.N. Allen, J.J. CorneveauX, D.A. Stephan, C.S. DeCarli, B.M. DeChairo,S.G. Potkin, C.R. Jack Jr., M.W. Weiner, C.A. Raji, O.L. Lopez, J.T. Becker, O.T. Carmichael, P.M. Thompson, I., Alzheimer’s disease neuroimaging, a commonly carried allele of the obesity-related FTO gene is associated with reduced brain volume in the healthy elderly, Proc. Natl. Acad. Sci. U S A 107 (18) (2010)8404–8409. [21] J. Li, L. Zeng, J. Xie, Z. Yue, H. Deng, X. Ma, C. Zheng, X. Wu, J. Luo, M. Liu, Inhibition of osteoclastogenesis and bone resorption in vitro and in vivo by a prenylflavonoid xanthohumol from hops, Sci. Rep. 5 (2015) 17605. [22] Y. Pan, P. Ma, Y. Liu, W. Li, Y. Shu, Multiple functions of m(6)A RNA methylation in cancer, J. Hematol. Oncol. 11 (1) (2018) 48. [23] D. Li, L. Cai, R. Meng, Z. Feng, Q. Xu, METTL3 modulates osteoclast differentiation and function by controlling RNA stability and nuclear export, Int. J. Mol. Sci. 21 (5) (2020). [24] Z. Li, H. Weng, R. Su, X. Weng, Z. Zuo, C. Li, H. Huang, S. Nachtergaele, L. Dong,C. Hu, X. Qin, L. Tang, Y. Wang, G.M. Hong, H. Huang, X. Wang, P. Chen,S. GurbuXani, S. Arnovitz, Y. Li, S. Li, J. Strong, M.B. Neilly, R.A. Larson, X. Jiang,P. Zhang, J. Jin, C. He, J. Chen, FTO plays an oncogenic role in acute myeloid leukemia as a N(6)-methyladenosine RNA demethylase, Cancer Cell 31 (1) (2017)127–141. [25] S. Bae, M.J. Lee, S.H. Mun, E.G. Giannopoulou, V. Yong-Gonzalez, J.R. Cross,K. Murata, V. Giguere, M. van der Meulen, K.H. Park-Min, MYC-dependentoXidative metabolism regulates osteoclastogenesis via nuclear receptor ERRalpha, J. Clin. Invest. 127 (7) (2017) 2555–2568. [26] L. Penolazzi, E. Lambertini, M. Borgatti, R. Piva, M. Cozzani, I. Giovannini,R. Naccari, G. Siciliani, R. Gambari, Decoy oligodeoXynucleotides targeting NF- kappaB transcription factors: induction of apoptosis in human primary osteoclasts, Biochem. Pharmacol. 66 (7) (2003) 1189–1198. [27] X. Gu, Y. Zhang, D. Li, H. Cai, L. Cai, Q. Xu, N6-methyladenosine demethylase FTO promotes M1 and M2 macrophage activation, Cell. Signal. 69 (2020), 109553. [28] M. Magnani, R. Crinelli, M. Bianchi, A. Antonelli, The ubiquitin-dependent proteolytic system and other potential targets for the modulation of nuclear factor-kB (NF-kB), Curr. Drug Targets 1 (4) (2000) 387–399. [29] T. Yamashita, Z. Yao, F. Li, Q. Zhang, I.R. Badell, E.M. Schwarz, S. Takeshita, E.F. Wagner, M. Noda, K. Matsuo, L. Xing, B.F. Boyce, NF-kappaB p50 and p52 regulate receptor activator of NF-kappaB ligand (RANKL) and tumor necrosis factor-induced osteoclast precursor differentiation by activating c-fos and NFATc1,J. Biol. Chem. 282 (25) (2007) 18245–18253. [30] H. Takatsuna, M. Asagiri, T. Kubota, K. Oka, T. Osada, C. Sugiyama, H. Saito,K. Aoki, K. Ohya, H. Takayanagi, K. Umezawa, Inhibition of RANKL-inducedosteoclastogenesis by (-)-DHMEQ, a novel NF-kappaB inhibitor, through downregulation of Dac51, J. Bone Miner. Res. 20 (4) (2005) 653–662.
[31] J.H. Park, N.K. Lee, S.Y. Lee, Current understanding of RANK signaling in osteoclast differentiation and maturation, Mol Cells 40 (10) (2017) 706–713.
[32] L.G. Raisz, Pathogenesis of osteoporosis: concepts, conflicts, and prospects, J. Clin. Invest. 115 (12) (2005) 3318–3325.