Inhibition of FAAH suppresses RANKL‐induced osteoclastogenesis and attenuates ovariectomy‐induced bone loss partially through repressing the IL17 pathway

Abstract

Fatty amide hydrolase (FAAH) is a key degradation enzyme of the endocannabinoid system, mainly responsible for the hydrolysis of arachidonic acid ethanolamine (AEA). Previous investigations have shown that FAAH is involved in a series of biological processes, such as inflammation, immune regulation, and transmembrane signal transduction of neurons. Endogenous cannabinoids and cannabinoid receptors have been reported to participate in the regulation of bone homeostasis by regulating the differentiation of osteoblasts and osteoclasts. We hypothesized that FAAH may play an important role in osteoclastogenesis based on the above evidence. The present study found that the FAAH expression was increased at both mRNA and protein levels during RANKL‐induced osteoclastogenesis. Pharmacological and genetic inhibition of FAAH in bone marrow‐derived macrophages (BMMs) inhibited osteoclastogenesis, F‐actin ring formation, bone resorption, and osteoclast‐specific gene expression in vitro. Moreover, intragastric administration of the FAAH inhibitor PF‐04457845(PF) ameliorated ovariectomy (OVX)‐induced bone loss in mice. Further investigation revealed that nuclear factor κB (NF‐κB) and mitogen‐activated protein kinase (MAPK) pathways were inhibited by PF treatment and FAAH knockdown. RNAseq indicated that the IL17 pathway was blocked by PF, and administration of recombinant murine IL17 protein could partially restore osteoclastogenesis and activate NF‐κB and MAPK pathways. To sum up, our findings demonstrate that targeting FAAH could be a promising candidate strategy for treating osteoclast‐related diseases, especially osteoporosis.

Abbreviations

2‐AG

arachidonic acid glycerol

AEA

arachidonic acid ethanolamine

BMMs

bone marrow‐derived macrophages

BV/TV

bone volume/tissue volume

CB

cannabinoid receptors

Ct. Ar

cortical bone area

Ct. Th

cortical thickness

CTSK

Cathepsin K

DMSO

dimethyl sulfoxide

FAAH

fatty amide hydrolase

FBS

fetal bovine serum

HE

hematoxylin and eosin

MAGL

monoacylglycerol lipase

MAPK

mitogen‐activated protein kinase

MCSF

macrophage colony‐stimulating factor

MMP9

matrix metalloproteinase 9

N. Oc/B. Pm

osteoclasts per bone perimeter

NFATc1

nuclear factor of activated T cells 1

NF‐κB

nuclear factor‐κB

OP

osteoporosis

OVX

ovariectomy

PBS

phosphate‐buffered saline

PCR

polymerase chain reaction

PF

PF‐04457845

RANKL

receptor activator for nuclear factor‐κB ligand

RIPA

radioimmune precipitation assay

RNAseq

RNA sequencing.

SDS‐PAGE

sodium dodecyl sulfate‐polyacrylamide gel electrophoresis

Tb. N

trabecular number

Tb. Sp

trabecular separation

Tb. Th

trabecular thickness

TRAF6

tumor necrosis factor receptor‐associated factor 6

TRAP

tartrate‐resistant acid phosphatase

α‐MEM

alpha modified minimal essential medium

μCT

microcomputed tomography

1. INTRODUCTION

Osteoporosis (OP) is a common metabolic bone disease affecting the skeletal system, characterized by decreased bone mineral density (BMD) and bone structural destruction, increasing the risk of fragility and fractures. ¹ , ² , ³ The elderly are much more likely to suffer from osteoporosis, particularly postmenopausal women. ⁴ It is estimated that around 200 million people worldwide suffer from the disease, and its prevention and treatment have become a common problem facing all mankind. ⁵ The bone is a dynamic organ, and its homeostasis is jointly maintained by osteoblastic bone formation and osteoclastic bone resorption. ⁶ Excessive bone resorption will lead to the imbalance of bone homeostasis and lead to a variety of orthopedic diseases, the most common of which is osteoporosis, Paget disease, rheumatoid arthritis, and periodontitis. ⁷ Although there are many kinds of drugs for the treatment of OP, such as estrogen, calcitonin, bisphosphate, teriparatide, and denosumab, which have certain curative effects, their limitations and side effects cannot be ignored, including osteonecrosis of the jaw, atypical femoral fracture, and thromboembolism. ⁸ , ⁹ Therefore, it is still urgent to further study the pathogenesis and explore new intervention targets and drugs for the prevention and treatment of osteoporosis.

Osteoclasts are multinucleated, giant cells differentiated from monocytes/macrophages under the stimulation of macrophage colony‐stimulating factor (MCSF) and nuclear factor receptor activator B (RANK) ligand (RANKL). They are also recognized as the main functional cells of bone resorption, possessing the ability to absorb and degrade the mineralized bone matrix. ¹⁰ M‐CSF is required for the survival and proliferation of monocyte/macrophage cell lines, while RANKL is crucial for osteoclast differentiation. ¹¹ , ¹² RANKL binds to its cell surface receptor RANK to recruit tumor necrosis factor (TNF) receptor‐associated factor 6 (TRAF6), thereby activating the downstream MAPK and NF‐κB pathways, and then induces the activation of transcription factors such as c‐Fos and nuclear factor‐activated T cell C1 (NFATc1), ultimately stimulating the expression of multiple osteoclast‐specific genes, such as tartrate‐resistant acid phosphatase (TRAP), matrix metallopeptidase 9 (MMP9) and cathepsin K (CTSK). ¹³ , ¹⁴

The endocannabinoid system consists of endogenous cannabinoids (such as AEA and arachidonic acid glycerol [2‐AG]), cannabinoid metabolic enzymes (such as FAAH and monoacylglycerol lipase [MAGL]), as well as cannabinoid receptor 1 and 2 (CB1 and CB2). ¹⁵ It plays a variety of biological roles in the body and participates in various diseases, such as inflammation, immune regulation, pain, tumor, energy homeostasis, and occurrence and development of cardiovascular disease. ¹⁶ , ¹⁷ , ¹⁸ FAAH is a key degradation enzyme of the cannabinoid system signal pathway, and its main physiological function is to degrade AEA into arachidonic acid and ethanolamine to terminate the action of AEA. ¹⁹ FAAH exerts a series of biological effects, such as inducing and antagonizing inflammatory response, regulating hematopoiesis and immunity, promoting cell division, regulating the transmembrane signal transduction of neurons and the release of neurotransmitters, thus participating in the pathophysiological processes of many diseases including myocardial infarction, obesity, and hyperlipidemia. ²⁰ , ²¹ , ²² However, the role of FAAH in osteoclastogenesis is still rarely reported. Notably, several studies have shown that the other two members of the endocannabinoid system, endocannabinoids and cannabinoid receptors, participate in the regulation of bone homeostasis by regulating the differentiation of osteoblasts and osteoclasts. AEA and 2‐AG have been proved to regulate bone formation, bone loss, bone tissue transformation, and other physiological phenomena. ²³ , ²⁴ , ²⁵ CB1 and CB2 are expressed in osteoblasts, osteoclasts, osteocytes, bone marrow mesenchymal stem cells (BMSC), and synoviocytes, and their activation play a complex role in regulating the differentiation of these cells, especially osteoblasts and osteoclasts. ²⁶ , ²⁷ , ²⁸ Given the above evidence, it is reasonable to speculate that FAAH may also be involved in osteoclast differentiation.

In this study, we investigated the role of FAAH in osteoclast differentiation in vitro and OVX‐induced osteoporosis by pharmacological and genetic inhibition of FAAH in a mouse model and explored the possible mechanisms.

2. MATERIALS AND METHODS

2.1. Reagents and antibodies

PF‐04457845 was purchased from Selleck (purity = 99.65%, Houston, USA). AM251 and AM630 were purchased from MedChemExpress (NJ, USA). They were diluted with DMSO (Sigma‐Aldrich, St. Louis, USA) and stored at −80°C. Recombinant mouse M‐CSF and RANKL were purchased from R&D Systems (Minneapolis, MN, USA). Recombinant mouse IL17 Protein was obtained from Absin (Shanghai, China). Cell Counting Kit‐8 (CCK‐8), phosphate buffer saline (PBS), and antibody against β‐actin were acquired from Boster Biological Technology (Wuhan, China). The TRAP staining kit was purchased from Sigma‐Aldrich (St. Louis, MO, USA). Phalloidin for F‐actin staining was purchased from the Beyotime Institute of Biotechnology (Jiangsu, China). The osteo assay surface for bone resorption was acquired from Corning Inc. Life Science (Corning, NY, USA). Specific antibodies against P65, p‐P65, IκBα, p‐IκBα, IKKβ, p‐IKKα/β, ERK, p‐ERK, JNK, p‐JNK, P38, p‐P38, NFTAc1, and c‐Fos were purchased from Cell Signaling Technology (Beverly, MA, USA). Primary antibodies specific for CTSK, MMP‐9, and TRAP were acquired from Abcam (Cambridge, MA, USA). Primary antibodies targeting FAAH, CB1, and CB2 were purchased from Abclonal (Wuhan, Hubei, China). The secondary antibodies were acquired from Jackson ImmunoResearch Laboratories (West Grove, PA, USA).

2.2. Cell culture and in vitro osteoclastogenesis assay

Primary bone marrow‐derived macrophages (BMMs) were isolated as previously described. ²⁹ Briefly, BMMs were isolated from the tibias and femurs of 6‐8‐week‐old male C57BL/6 mice and cultured in α‐MEM containing 10% fetal bovine serum (FBS) and 30 ng/ml M‐CSF for 16 h. No ‐ adherent cells were transferred to another 10 cm dish with M‐CSF supplementation and cultured for an additional 3 days. The attached cells were then digested with 0.25% trypsin for experimental purposes. For osteoclastogenesis assay, BMMs were seeded in 96‐well plates at a density of 2 × 10⁴ cells per well in triplicate and cultured in osteoclastogenic medium containing M‐CSF (30 ng/ml) and RANKL (75 ng/ml) for 5–7 days to differentiate into osteoclasts. Multinucleated osteoclasts were then fixed and stained using the TRAP kit according to the manufacturer's protocol. TRAP‐positive multinucleated cells with 3 or more nuclei were identified as mature osteoclasts.

2.3. Cell proliferation assay

The CCK‐8 assay was utilized following the manufacturer's instructions to determine the effects of PF on BMMs viability. In brief, BMMs were seeded in 5 identical 96‐well plates (8 × 10³ cells per well), incubated with 30 ng/ml M‐CSF overnight, and then treated with different concentrations of PF (0, 0.25, 0.5, 1, 2.5, 5, and 10 μM) for 1, 3 and 5 days. Thereafter, the BMMs were incubated upon addition of 10 μl of CCK‐8 solution in each well for 1 h at 37°C. Finally, absorbance was measured at a wavelength of 450 nm by a microplate reader (Bio‐Tek).

2.4. Lentivirus transfection

For knock‐down of FAAH expression, lentiviral vectors encoding shRNA against murine FAAH and control lentivirus were acquired from Shanghai GeneChem Company. The shRNA sequences were as follows: shRNA1, 5′‐CCGGGCCCAGATGGAACACTACAAACTCGAGTTTGTAGTGTTCCATCTGGGCTTTTTG‐3′, shRNA2, 5′‐CCGGCCCTTCTTACCAAACAACATACTCGAGTATGTTGTTTGGTAAGAAGGGTTTTTG‐3′, shRNA3, 5′‐CCGGGCATTGTGCATGAAAGCCCTACTCGAGTAGGGCTTTCATGCACAATGCTTTTTG‐3′. For lentivirus infection, BMMs were cultured with lentiviruses in the α‐MEM medium for 24 h. Knockdown effects were determined by qPCR and Western blot.

2.5. F‐actin staining and pit formation assay

F‐actin ring staining and pit formation assays were performed as described earlier to analyze the effect of FAAH inhibition on osteoclast function. ³⁰ Briefly, BMMs were cultured on 0.2% collagen‐gel coated 6‐well plates with M‐CSF (30 ng/ml) and RANKL (75 ng/ml) until osteoclasts formed. Subsequently, RANKL‐induced mature osteoclasts were digested with type I collagenase (0.2%) and reseeded in Corning osteo assay strip wells, and stimulated with different concentrations of PF (0, 0.25, 0.5, 1, and 2.5 μM) or infected with the lentivirus carrying FAAH‐specific shRNA or the control lentivirus for an additional 3 days in the presence of 75 ng/ml RANKL. For F‐actin staining, the cells were fixed in 4% paraformaldehyde for 15 min and permeabilized with 0.2% Triton X‐100 (Solarbio) for 5 min, then stained with Phalloidin labeled with rhodamine for 1 h at room temperature and counterstained with DAPI for 5 min. Actin rings were captured under a fluorescence microscope and analyzed using ImageJ software (National Institutes of Health, Bethesda, MD, USA). For pit formation assay, the cells were washed with 5% sodium hypochlorite for 5 min and resorption pits were captured under light microscopy and analyzed using ImageJ software.

2.6. Quantitative real‐time PCR (qPCR)

Total RNA was extracted with TRIzol reagent (Invitrogen), and cDNA was reverse transcribed from 1 μg of total RNA using the Hieff First Strand cDNA Synthesis SuperMix (Yeasen Biotechnology) based on the manufacturer's specifications. The Hief qPCR SYBR Green Master Mix (Yeasen Biotechnology) was used for qPCR, which proceeded using the following parameters: 95°C for 5 min, followed by 95°C (10 s) and 60°C (30 s) for 40 cycles. The relative level of target genes was normalized to the housekeeping gene GAPDH. The primer sequences used were listed in Table S1.

2.7. Western blot analysis

Total protein was extracted with RIPA lysis buffer containing 1% proteinase inhibitor and phosphotransferase inhibitor (Boster Biotechnology). The proteins were separated by 10% sodium dodecyl sulfate‐polyacrylamide gel electrophoresis and then transferred onto PVDF membranes (Millipore, Billerica, USA), which were blocked with 5% nonfat milk in TBST for 1 h. Thereafter, the PVDF membranes were incubated with respective primary antibodies under continuous shaking overnight at 4°C. Next, the PVDF membranes were washed three times with PBS and incubated with secondary antibodies (1: 10000) for 1 h at room temperature. Finally, the immunoreactive proteins were visualized with extremely sensitive chemiluminescence (Thermo Fisher Scientific) and analyzed by the Image Lab 5.1 software (Bio‐Rad, Hercules, CA, USA).

2.8. Immunofluorescence analysis

Firstly, the treated cells were fixed in 4% paraformaldehyde for 15 min and permeabilized with 0.2% Triton X‐100 (Solarbio) for 5 min. The nonspecific binding of the first antibody was blocked by 1% BSA in PBST at room temperature for 1 h. Subsequently, the cells were incubated with diluted primary antibodies under continuous shaking at 4°C overnight. The next day, the cells were washed with PBST three times and incubated with appropriate secondary antibodies in the dark for 1 h, followed by incubation with DAPI (Boster Biological Technology) for 5 min. Finally, images were captured using a fluorescence microscope.

2.9. RNA sequencing

For RNA‐sequencing (RNAseq), BMMs were treated with PF (2.5 μM) and stimulated with 75 g/ml RANKL for 3 days. Total RNA was extracted using TRIzol reagent (Invitrogen), and RNA quality was determined using 2100 Bioanalyser (Agilent) and quantified using the ND‐2000 (NanoDrop Technologies). The subsequent RNA‐sequencing analysis was performed by Illumina NovaSeq 6000 sequencing (150 bp*2, Shanghai BIOZERON Co., Ltd). The clean reads were separately aligned to reference genome with orientation mode using hisat2 (https://ccb.jhu.edu/software/hisat2/index.shtml) software. Fragments per kilobase of exon per million mapped reads (FRKM) method was used to identify differential expression genes (DEGs) between the two different samples. The false discovery rate (FDR) < 0.05 and the logarithmic foldchange >2 were set as the threshold for significantly differential expression. To investigate the biological functions of the differentially expressed genes, GO functional enrichment and KEGG pathway analysis were performed using Goatools (https://github.com/tanghaibao/Goatools) and KOBAS (http://kobas.cbi.pku.edu.cn/home.do).

2.10. Osteoblastogenesis assay

MC3T3‐E1 cells were seeded in 24‐well plates at a density of 5 × 10⁴ cells/well in triplicate and incubated in osteogenic differentiation medium containing 5 mM β‐glycerophosphate, 100 nM dexamethasone, and 50 μg/ml ascorbic acid with different concentrations of PF (0, 0.5, 1, and 2.5 μM). After 7‐day and 21‐day osteogenic induction, cells were fixed in 4% paraformaldehyde and then stained with alkaline phosphatase (ALP) solution and Alizarin Red S solution (Cyagen Biosciences) respectively according to the manufacturer's instructions. Images were captured under light microscopy.

2.11. Murine model of OVX‐induced bone loss

All animal experimental surgeries were conducted following National Institutes of Health (NIH) guidelines for the care and use of laboratory animals and approved by Tongji Hospital Institutional Animal Ethics Committee. Thirty‐two C57BL/6 mice (female, 12 weeks old) were acquired from the Experimental Animal Center of Tongji Hospital (Wuhan, China) and housed under standard living conditions for 1 week. The mice were randomly segregated into four groups (8 mice per group) based on a random number table: SHAM group (sham‐surgery + vehicle treatment), OVX group (surgically OVX + vehicle treatment), OVX with low‐dose PF group (OVX + 1 mg/kg PF, intragastric administration), and OVX with high‐dose PF group (OVX + 10 mg/kg PF, intragastric administration). OVX and sham surgeries were performed after a week of adaptive feeding as previously described. ³¹ Briefly, ovaries were exteriorized and removed via bilateral dorsal approach under pentobarbital anesthesia, and the sham surgeries were performed by only exposing the bilateral ovaries. One week after the operation, all surgical mice were treated 5 days per week by feeding PF or vehicle for 8 consecutive weeks. In the end, mice were euthanized to collect sera, femurs, and vertebras for further analysis.

2.12. Micro‐computed tomography (μCT) scanning and histomorphometric analysis

The harvested femurs from mice were fixed in 4% paraformaldehyde for 3 days. Then, the left distal femurs were scanned with high‐resolution μCT (Scanco Medical, Bassersdorf, Switzerland). The parameters were as follows: source voltage, 100 kV; source current, 98 μA; voxel size, 10 μm; and exposure time, 300 ms. The three‐dimensional images were reconstructed and analyzed by using the built‐in μCT software with the following trabecular parameters: bone volume/tissue volume (BV/TV), trabecular number (Tb. N), trabecular thickness (Tb. Th), and trabecular separation (Tb. Sp). The region of interest selected for analysis was 5% of the femoral length from 0.1 mm below the growth plate. ³² For histomorphometric analysis, the fixed right femurs were decalcified in 10% ethylenediamine tetraacetic acid (EDTA, Sigma‐Aldrich) for 3 weeks. Then, the paraffin‐embedded femurs were sectioned into slices at a 5‐μm thickness for hematoxylin and eosin (H&E) staining, TRAP staining, and immunohistochemical staining. Stained sections were visualized using an axiovert optical microscope.

2.13. ELISA

Mouse serum samples were collected and stored at −80°C until analyzed. Serum IL17 and P1NP were measured using IL17 ELISA Kit (ABclonal Technology Co., Ltd.) and P1NP ELISA Kit (Cloud‐Clone Corp, Houston, USA) according to the manufacturer's protocols.

2.14. Statistical analysis

Quantitative data were presented as mean ± SD of at least three independent experiments. Statistical analyses were performed using Prism 8 software (GraphPad Software Inc., La Jolla, CA, USA). Statistical differences between two groups were determined with unpaired two‐tailed Student's t‐test, and multiple comparisons were assessed using one‐way ANOVA followed by appropriate Tukey's post hoc analysis. Differences with p < .05 were considered statistically significant.

3. RESULTS

3.1. Inhibition of FAAH represses osteoclast formation and bone resorption without significant cytotoxicity in vitro

To identify the role of FAAH in osteoclastogenesis, we first detected the expression of FAAH during RANKL‐induced osteoclast differentiation. RNA and protein were extracted at 0, 1, 3, and 5 days after osteoclast differentiation induced by RANKL, and detected with qPCR and Western blot assays, respectively. The results showed that FAAH expression at both mRNA and protein levels was significantly increased during osteoclastogenesis (Figure 1A–C), which was also confirmed by the immunofluorescence experiment (Figure 1D). To further clarify the role of FAAH in osteoclast formation and function, BMMs were infected with lentivirus carrying FAAH‐specific shRNA or control lentivirus to down‐regulated the expression of FAAH. Gene knockout efficiency was detected using qPCR and Western blot assays (Figure 1E–G). The results of TRAP staining showed that compared with the control group, shRNA with the highest gene knockout efficiency had a significant inhibitory effect on osteoclast formation (Figure 1H,I). To further explore the effect of FAAH on the function of osteoclasts, we carried out F‐actin ring staining and pit formation assays, which were consistent with TRAP staining, indicating that FAAH‐shRNA lentivirus could significantly inhibit F‐actin ring formation (Figure 1J,K) and osteoclast bone resorption (Figure 1L,M).

FIGURE 1.

FAAH is up‐regulated during osteoclast differentiation, and the knockdown of FAAH suppresses osteoclast formation and bone resorption. (A–C) BMMs were cultured with 30 ng/ml M‐CSF and 75 ng/ml RANKL for 0, 1, 3, and 5 days. The changes in mRNA and protein levels of FAAH during osteoclast differentiation were detected by qPCR and Western blot respectively. (D) BMMs were cultured with 30 ng/ml M‐CSF and 75 ng/ml RANKL for 0, 1, 3, and 5 days, then the cells were fixed with 4% paraformaldehyde for immunofluorescence analysis. Scale bars = 200 μm. (E–G) BMMs were transfected with three lentiviruses carrying FAAH shRNA or a control lentivirus, and the knockdown efficiency of FAAH was measured by qPCR and Western blot. (H and I) BMMs were infected with the lentivirus carrying FAAH‐specific shRNA or the control lentivirus and then treated with RANKL for 5–7 d. TRAP staining was performed, and multinucleated (>3 nuclei) cells were quantified. Scale bars = 400 μm. (J–M) RANKL‐induced mature osteoclasts were seeded in Corning osteo assay strip wells and infected with the lentivirus carrying FAAH‐specific shRNA or the control lentivirus, then stimulated with RANKL for an additional 3 days. F‐actin staining or pit formation assays were performed. Scale bars = 400 μm. All data are presented as the mean ± SD. *p < .05, **p < .01, ns, non‐significant, compared with the control groups. N = 3 per group.

In addition, PF, a selective inhibitor of FAAH, was used to further evaluate the effect of inhibiting FAAH on osteoclast formation and function. Firstly, a CCK8 assay was conducted to detect the effect of PF on BMMs activity. The results showed that different concentrations of PF (0, 0.25, 0.5, 1, 2.5, 5, and 10 μM) had no significant toxicity to BMMs viability at different times (1, 3, and 5 days) (Figure 2A). Subsequently, BMMs were treated with RANKL to induce osteoclast differentiation at different concentrations of PF (0, 0.25, 0.5, 1, and 2.5 μM). The results of TRAP staining were consistent with the knockout effect of FAAH‐shRNA, which also showed that PF could significantly inhibit osteoclast formation in a concentration‐dependent manner (Figure 2B,C). To further determine the main stage of PF inhibition on osteoclast differentiation, 2.5 μM PF was given at different stages of osteoclast differentiation. TRAP staining showed that PF mainly inhibited osteoclast formation in the early stage (Figure 2D,E). Similarly, F‐actin ring staining and pit formation assays showed that PF could significantly impair F‐actin ring formation (Figure 2F,G) and inhibit osteoclast bone resorption (Figure 2H,I) in a concentration‐dependent manner.

FIGURE 2.

The FAAH antagonist PF‐04457845 (PF) represses osteoclast formation and bone resorption without significant cytotoxicity. (A) BMMs were treated with M‐CSF (30 ng/ml) and various concentrations of PF (0, 0.25, 0.5, 1, 2.5, 5, and 10 μM) for 1, 3, and 5 days. Cell viability was assessed by a CCK8 assay. (B and C) BMMs were treated with different concentrations of PF (0, 0.25, 0.5, 1, and 2.5 μM) in the presence of 75 ng/ml RANKL for 5–7 days. (D and E) BMM were cultured in the presence of 30 ng/ml M‐CSF and 75 ng/mL RANKL, and were treated with 2.5 μM PF starting at the indicated time of osteoclastogenesis until osteoclasts were fully formed in the control group. TRAP staining was performed, and multinucleated (>3 nuclei) osteoclasts were quantified. (F–I) RANKL‐induced mature osteoclasts were digested and seeded in Corning osteo assay strip wells and treated with different concentrations of PF (0, 0.25, 0.5, 1, and 2.5 μM) for an additional 3 days in the presence of 75 ng/ml RANKL. F‐actin staining (F and G) and pit formation assays (H and I) were conducted. All data are presented as the mean ± SD. *p < .05, **p < .01, ns, non‐significant, compared with the control groups. N = 3 per group. Scale bars = 400 μm.

3.2. Inhibition of FAAH represses osteoclast‐specific genes expression during osteoclastogenesis

To examine whether the inhibition of FAAH affected the expression levels of osteoclast‐specific genes, BMMs were first treated with lentivirus carrying FAAH‐shRNA and induced with RANKL for 3 days. qPCR and Western blot confirmed that the expression levels of NFATc1, c‐Fos, TRAP, MMP9, and CTSK were significantly increased after RANKL stimulation, while the knockdown of FAAH significantly decreased their expression levels (Figure 3A–C). Subsequently, BMMs were treated with RANKL to induce osteoclast differentiation at different concentrations of PF (0, 0.5, 1, and 2.5 μM), the results of qPCR and Western blot were consistent with the knockout effect of FAAH‐shRNA, and the inhibitory effect of PF was dose‐dependent (Figure 3D–F). At the same time, BMMs were treated with or without 2.5 μM PF in the presence of 75 ng/ml RANKL for 0, 1, 3, and 5 days, the results of qPCR and Western blot confirmed that PF inhibited the expression of specific genes in the process of osteoclast formation, and the inhibitory effect was time‐dependent (Figure 3G–I).

FIGURE 3.

Inhibition of FAAH represses osteoclast‐specific gene expression during osteoclastogenesis. (A–C) BMMs were infected with the lentivirus carrying FAAH‐specific shRNA or the control lentivirus and then treated with RANKL for 3 days, and the mRNA and protein expression of FAAH and osteoclast‐specific genes were detected by qPCR and Western blot respectively. *p < .05, **p < .01. (D–F) BMMs were treated with indicated concentrations of PF (0, 0.5, 1, and 2.5 μM) with or without stimulation by 75 ng/ml RANKL for 3 days, and the mRNA and protein expression of osteoclast‐specific genes were determined using qPCR and Western blot respectively. ^# p < .05, ^## p < .01, compared with the control groups; *p < .05, **p < .01, compared with the RANKL groups. (G–I) BMMs were treated with or without 2.5 μM PF in the presence of 75 ng/ml RANKL for 0, 1, 3, and 5 days. mRNA and protein levels of osteoclast‐specific genes were measured using qPCR and Western blot respectively. *p < .05, **p < .01, compared with the control groups at the same time point. All data are presented as the mean ± SD. N = 3 per group.

3.3. Inhibition of FAAH has no significant effect on osteogenesis and osteoblast‐related gene expression

It is known that bone homeostasis is maintained by osteoblastic bone formation and osteoclastic bone resorption. Therefore, after confirming the role of FAAH in osteoclast differentiation, we also preliminarily explored its role in osteoblast differentiation with MC3T3‐E1 cells. Firstly, the CCK8 results showed that different concentrations of PF (0, 0.25, 0.5, 1, 2.5, 5, and 10 μM) had no significant effect on the activity of MC3T3‐E1 cells after different days of intervention (2, 4, 6 days) (Figure 4A). Subsequently, osteogenic differentiation was induced under the intervention of different concentrations of PF (0, 0.5, 1, and 2.5 μM). ALP staining and ARS staining showed that PF intervention had no significant effect on osteogenic differentiation and mineralization (Figure 4B). In addition, qPCR and Western blot assays also confirmed that PF treatment had no significant effect on the expression of osteoblast marker genes ALP, Runx2, OPN, and OCN at both the mRNA and protein levels (Figure 4C–E). Moreover, the immunohistochemical staining of bone tissue sections and serum ELISA showed that PF treatment had no significant improvement on the decrease of bone formation marker COL1 in bone and P1NP in serum induced by OVX (Figure S1).

FIGURE 4.

Inhibition of FAAH has no significant effect on osteogenesis and osteoblast‐related gene expression. (A) MC3T3‐E1 cells were treated with various concentrations of PF (0, 0.25, 0.5, 1, 2.5, 5, and 10 μM) for 2, 4, and 6 days. Cell viability was assessed by a CCK8 assay. (B) MC3T3‐E1 cells were cultured in osteogenic induction medium with different concentrations of PF (0, 0.5, 1, and 2.5 μM). ALP staining and ARS staining were performed on days 7 and 21, respectively. (C–E) MC3T3‐E1 cells were treated with indicated concentrations of PF (0, 0.5, 1, and 2.5 μM) under osteogenic induction, and the mRNA and protein expression of osteoblast‐specific genes were determined using qPCR and Western blot on days 7 and 14, respectively. All data are presented as the mean ± SD. ns, non‐significant, compared with the control groups. N = 3 per group.

3.4. PF administration ameliorates OVX‐induced bone loss

The results of in vitro experiments showed that inhibition of FAAH had a significant inhibitory effect on osteoclastogenesis and bone resorption. On this basis, we further conducted animal experiments to simulate postmenopausal osteoporosis using a mouse OVX model and assessed the ability of intragastric administration of specific FAAH inhibitor PF to prevent OVX‐induced osteoporosis in vivo. One week after surgery, the mice were given intragastric administration of 1, 10 mg/kg PF, or vehicle for 8 weeks, at least 5 times a week. Thereafter, the mice were sacrificed, and the femurs and vertebrae were collected for μCT scanning and bone tissue section staining (Figure 5A). During the experiment, the mice were weighed once a week, and it was found that intragastric administration of PF did not inhibit the weight gain of mice, reaffirming its safety to some extent (Figure 5B). The results of the μCT scan showed that the bone trabeculae in the OVX group were significantly less than the SHAM group, which confirmed the success of the OVX model. However, compared with the OVX group, the trabecular loss was significantly reduced in the OVX + PF group (Figure 5C). In addition, reconstruction results showed that compared with the OVX group, the related bone parameters BV/TV, Tb. N, and Tb. Th of the OVX + PF group was significantly increased, Tb. Sp was significantly decreased, with a higher dose of PF (10 mg/kg) showing a better effect on improving bone loss (Figure 5D). It's worth noting that there was no significant difference between the OVX + PF10 group and the SHAM group, indicating that the bone trabeculae of the OVX + PF10 group had basically returned to the level of the SHAM group. Similarly, μCT scan and reconstruction of the 5th lumbar vertebrae confirmed the effect of PF on trabecular loss of axial bone induced by OVX in mice, although the results were not as significant as that of femur scan reconstruction (Figure 5E,F). The cortical bone from the femoral midshaft was also scanned and reconstructed, and the related bone parameters Ct. Th and Ct. Ar were analyzed, but no significant change was found among the groups (Figure 5G,H).

FIGURE 5.

PF administration ameliorates OVX‐induced bone loss in vivo. (A) Schematic diagram of the experiments in vivo. (B) The weight of postoperative mice was weighed once a week. (C and E) Representative three‐dimensional reconstructed μCT images of distal femurs and the 5th lumbar vertebrae (coronal and axial planes) of the mice were shown. (D and F) Quantitative analyses of bone structural parameters of the distal femurs and the 5th lumbar vertebrae, including BV/TV, Tb. N, Tb. Th, and Tb. Sp. (G and H) Representative three‐dimensional reconstructed μCT images of cortical bone from the femoral midshaft and quantitative analyses of cortical bone parameters, including Ct. Th and Ct. Ar. (I) Representative sections of the distal femurs were performed with H&E staining (scale bar = 1000 μm) and TRAP staining (scale bar = 400 μm). (J) Quantitative analyses of histomorphometric bone parameters, including N. Oc/B. Pm and Oc. S/BS were performed. All data are presented as the mean ± SD. *p < .05, **p < .01. N = 6–8 per group.

In addition, the histological staining also confirmed the protective effect of PF on bone loss induced by OVX. H&E staining showed that trabecular loss was significant reduced in the OVX + PF group compared with the OVX group (Figure 5I). TRAP staining showed that the number of osteoclasts per bone perimeter (N. Oc/B. Pm) of the distal femur in the OVX group was significantly higher than that in the SHAM group, and osteoclast surface per bone surface (Oc. S/BS) was increased, but they were significantly decreased by administration of PF, although it had not yet returned to the level of SHAM group (Figure 5I,J). Besides, IHC analysis showed that the expression of FAAH in bone tissue of the OVX group was significantly higher than that of the SHAM group (Figure S2). In conclusion, inhibition of FAAH with selective inhibitor PF can prevent OVX‐induced bone loss by inhibiting osteoclasts activity in vivo.

3.5. PF inhibits osteoclast differentiation – CB1 or CB2 independence

Previous studies have confirmed that both cannabinoid receptors CB1 and CB2, play important roles in bone homeostasis. Therefore, we first thought about whether the regulation of FAAH involved in osteoclast differentiation was mediated by classical CB1 and CB2 receptors in the classical endocannabinoid system. Firstly, we used lentivirus encoding FAAH‐shRNA and PF to intervene BMMs, and then detected the expression of CB1 and CB2 during the differentiation of BMMs into osteoclast by qPCR and Western blot assays, respectively. The results showed that neither gene knockdown of FAAH (Figure 6A–C) nor pharmacological inhibition (Figure 6D–F) had a significant effect on the expression of CB1 and CB2. Subsequently, selective receptor antagonists AM251 (0.1 μM) and AM630 (0.1 μM) were used to selectively block CB1 and CB2 receptors respectively for rescue experiments. TRAP staining results showed that neither the antagonists AM251 nor AM630 could rescue the inhibitory effect of PF on osteoclast formation (Figure 6G,H), which was also verified by qPCR and Western blot assays at mRNA and protein levels, respectively (Figure 6I–K). These results suggested that the inhibitory effect of FAAH inhibition on osteoclast differentiation was independent of the activation of CB1 and CB2 receptors.

FIGURE 6.

PF inhibits osteoclast differentiation – CB1 or CB2 independence. (A–C) BMMs were infected with the lentivirus carrying FAAH‐specific shRNA or the control lentivirus and then treated with RANKL for 3 days, and the mRNA and protein expression of CB1 and CB2 were detected by qPCR and Western blot respectively. (D–F) BMMs were treated with indicated concentrations of PF (0, 0.5, 1, and 2.5 μM) with or without stimulation by 75 ng/ml RANKL for 3 days, and the mRNA and protein expression of CB1 or CB2 were determined using qPCR and Western blot respectively. ^# p < .05, ^## p < .01, compared with the control groups; ns, non‐significant, compared with the RANKL groups. (G and H) BMMs were treated with PF (2.5 μM), AM251 (0.1 μM), and AM630 (0.1 μM) as indicated in the presence of 75 ng/ml RANKL. After 5–7 d, trap staining was conducted, and multinucleated (>3 nuclei) cells were quantified (scale bar = 400 μm). (I–K) BMMs were treated with PF (2.5 μM), AM251 (0.1 μM), and AM630 (0.1 μM) as indicated in the presence of 75 ng/ml RANKL for 3 days, and the mRNA and protein expression of osteoclast‐specific genes were determined using qPCR and Western blot respectively. ^# p < .05, ^## p < .01, compared with the RANKL groups; ns, non‐significant, compared with the RANKL + PF groups. All data are presented as the mean ± SD. N = 3 per group.

3.6. RNA sequencing indicates multiple signaling pathways as downstream factors of FAAH inhibitor during osteoclast differentiation

To further elucidate the mechanism of the effect of FAAH on osteoclast differentiation, we performed the whole genome RNA sequencing (RNAseq) analysis on primary osteoclast differentiation with or without PF intervention and compared the two groups to explore the downstream genes and signal pathways that may be affected by FAAH inhibition. Gene alteration with a fold change greater than 2 and an adjusted P value less than 0.05 was assigned as differentially expressed. The heat map showed the top 100 differentially expressed genes with the greatest multiple changes after 3 days of RANKL and PF stimulation (Figure 7A), many of which were marker genes of osteoclast differentiation induced by RANKL, suggesting that FAAH was involved in the process of RANKL‐mediated osteoclast differentiation. KEGG and GSEA enrichment analyses were used to determine the top differentially regulated pathways. KEGG pathway analysis showed that after the intervention of PF, IL17 signaling pathway, Osteoclast differentiation, TNF signaling pathway, Fc gamma R‐mediated phagocytosis, NF‐κB signaling pathway, Regulation of actin cytoskeleton, and NOD‐like receptor signaling pathway were significantly down‐regulated, among which the first two pathways with the most significant changes were the IL17 signaling pathway and osteoclast differentiation pathway, respectively (Figure 7B,C). The above results of the KEGG enrichment analysis were further confirmed by GSEA analyses (Figure 7D). Several genes in the above IL17 signal pathway, including Il17ra, Act1, Lcn2, Il1 β, S100a8, and S100a9, were verified to be inhibited by PF during osteoclast differentiation with qPCR assay, which was consistent with RNAseq (Figure 7E). These results suggested that the IL17 signal pathway is a key mediator of FAAH on osteoclast differentiation.

FIGURE 7.

RNA sequencing indicates multiple signaling pathways as downstream factors of FAAH inhibitor during osteoclast differentiation. (A) The heat map showed the top 100 differentially expressed genes with the greatest multiple changes in response to PF stimulation. (B and C) KEGG enrichment analysis showed the most varied signal pathways and the number of genes contained in these pathways after PF intervention, among which the first two pathways with the most significant changes were the IL17 signaling pathway and osteoclast differentiation pathway, respectively. (D) GSEA analysis further confirmed the results of the KEGG enrichment analysis. The IL17 signaling pathway and osteoclast differentiation pathway were changed significantly after PF intervention. (E) Several genes in the above IL17 signal pathway, including Il17ra, Act1, Lcn2, Il1 β, S100a8, and S100a9, were verified by qPCR assay. All data are presented as the mean ± SD. **p < .01, compared with the RANKL groups. N = 3 per group.

3.7. Inhibition of FAAH suppresses RANKL‐induced NF‐κB and MAPK signaling activation

Previous studies have shown that NF‐κB and MAPK pathways play a crucial role in RANKL‐induced osteoclast differentiation. KEGG enrichment analysis also found that these two signal pathways were significantly inhibited after PF intervention. Therefore, we further verified the effect of FAAH inhibition on NF‐κB and MAPK signal pathways at the protein level. We used lentivirus carrying FAAH‐shRNA and 2.5 μM PF intervention to inhibit the expression of FAAH and treated BMMs with RANKL (75 ng/ml) for 0, 15, 30, and 60 min, respectively. Western blot showed that both gene knockout (Figure 8A,C) and pharmacological inhibition (Figure 8B,D) could inhibit RANKL‐induced phosphorylation of IKKα/β, IκBα, and p65 in NF‐κB pathway. At the same time, both gene knockout (Figure 8E,G) and pharmacological inhibition (Figure 8F,H) could inhibit the phosphorylation of ERK, p38, and JNK in MAPK pathway. Although FAAH knockout and PF treatment had different effects on NF‐κB and MAPK pathway family members, both of them have been shown to inhibit the activation of NF‐κB and MAPK signal pathways.

FIGURE 8.

Inhibition of FAAH suppresses RANKL‐induced NF‐κB and MAPK signaling activation. (A, C, E, and G) BMMs were infected with the lentivirus carrying FAAH‐specific shRNA or the control lentivirus for 2 days and then starved in the medium without FBS for 16 h. Subsequently, the cells were stimulated with RANKL (75 ng/ml) for 0, 15, 30, and 60 min. Finally, Western blot analysis was used to analyze the expression of the total and phosphorylated protein of NF‐κB and MAPK signaling pathways. (B, D, F, and H) BMMs were starved for 16 h in the presence or absence of 2.5 μM PF and then stimulated with 75 ng/ml RANKL with or without 2.5 μM PF for the indicated times. The total and phosphorylated protein expression levels of NF‐κB and MAPK signaling pathways were analyzed using Western blot assay. All data are presented as the mean ± SD. *p < .05, **p < .01, compared with the control groups at the same time point. N = 3 per group.

3.8. IL17 signal pathway is a key mediator of FAAH on osteoclast differentiation

To further confirm the role of the IL17 signal pathway in FAAH‐mediated osteoclast formation and functional regulation, we used recombinant murine IL17 protein to interfere with PF‐mediated BMMs. The results showed that IL17 could partially restore the inhibitory effect of PF on osteoclast formation (Figure 9A,C), F‐actin ring formation (Figure 9B,F), and bone resorption of osteoclasts (Figure 9D,E). In addition, we also verified whether IL17 could rescue the inhibitory effect of PF on the osteoclast marker gene and RANKL signal pathway. qPCR analysis confirmed that inhibition of FAAH with PF could inhibit the expression of NFATc1, MMP9, TRAP, and CTSK at the mRNA level, while supplementation of IL17 partially restored their expression (Figure 9G). The salvage effect of IL17 on the osteoclast marker gene was also confirmed at the protein level by Western blot assay (Figure 9H,I). Then, we examined the effects of IL17 on NF‐κB and MAPK signal pathways after PF intervention. Western blot assay confirmed that IL17 could partially reverse the inhibitory effect of PF on NF‐κB pathway (Figure 9J,L) and MAPK pathway (Figure 9K,M). The results of animal experiments were consistent with cell experiments. Immunohistochemical staining of IL17 showed that the expression of IL17 increased significantly after OVX, while the effect of PF could notably reduce its expression (Figure S3A,B), which was also confirmed by ELISA (Figure S3C).

FIGURE 9.

IL17 Signaling activation partially alleviates PF‐mediated inhibitory effects of osteoclast differentiation and the activation of NF‐κB and MAPK signaling pathways. (A–F) BMMs were treated with PF (2.5 μM) and recombinant murine IL17 protein (1 ng/ml) as indicated in the presence of 75 ng/ml RANKL. After 5–7 days, TRAP staining (A and C), F‐actin staining (B and F), and pit formation assays (D and E) were conducted and quantified. **p < .01, compared with the control groups; ^## p < .01, compared with the PF groups. scale bar = 400 μm. (G–I) BMMs were treated with PF (2.5 μM) and recombinant murine IL17 protein (1 ng/ml) as indicated in the presence of 75 ng/ml RANKL for 3 days, and the mRNA and protein expression of osteoclast‐specific genes were determined using qPCR and Western blot respectively. *p < .05, **p < .01, compared with the control groups; ^# p < .05, ^## p < .01, compared with the PF groups. (J–M) BMMs were starved for 16 h with or without 1 ng/ml recombinant murine IL17 protein and then stimulated with RANKL (75 ng/ml) in the presence or absence of 1 ng/ml recombinant murine IL17 protein for 0, 15, 30, and 60 min. The total and phosphorylated protein expression levels of NF‐κB and MAPK signaling pathways were analyzed by Western blot. *p < .05, **p < .01, compared with the RANKL groups at the same time point. All data are presented as the mean ± SD. N = 3 per group.

4. DISCUSSION

Osteoporosis is a common metabolic bone disease characterized by a deregulated coupling between bone formation and bone resorption, leading to decreased bone mass and the destruction of bone microarchitecture. ³³ The coupling disorder of bone formation and bone resorption is mostly caused by excessive bone resorption due to increased osteoclast formation. ³⁴ Therefore, inhibition of progressive osteoclast differentiation is an effective therapeutic strategy for osteoporosis and other osteoclast‐related diseases. In the present study, we demonstrated the role of FAAH in osteoclastogenesis. We found that the expression of FAAH was enhanced during osteoclastogenesis, suggesting that FAAH may play a positive role in osteoclast osteoclastogenesis. Lentivirus carrying FAAH‐shRNA and PF‐04457845 (PF), both showed inhibitory effects on osteoclastogenesis, bone resorption, the expression of osteoclast marker genes, and the activation of MAPK and NF‐κB pathways. However, FAAH inhibition had no significant effect on osteoblast differentiation and the expression of osteoblast‐specific genes. In vivo experiments showed that pharmacological inhibition of FAAH by PF ameliorated OVX‐induced bone loss and even restored the trabecula to the physiological level. These results suggested that FAAH may be an effective target for the treatment of osteoclast‐related diseases.

FAAH is a key enzyme regulating endocannabinoid signaling transduction, mainly responsible for the hydrolysis of AEA, and plays a series of biological effects such as inducing and antagonizing inflammatory responses, regulating hematopoiesis and immunity, promoting cell division, and regulating neuronal transmembrane signaling transduction. ³⁵ , ³⁶ Inhibitors of FAAH can exert therapeutic effects without causing the adverse reactions of THC directly stimulating cannabinoid CB1 receptors. ¹⁸ , ³⁷ A previous study showed that pharmacological inactivation of FAAH in monocyte‐macrophages from AD patients reduced the production of pro‐inflammatory cytokines IL‐6, IL‐12, and TNF‐α. ³⁸ Many studies have confirmed that FAAH inhibitor URB597 has analgesic, anxiolytic, and antidepressant effects in vivo, as well as certain efficacy against epilepsy and schizophrenia. ³⁹ , ⁴⁰ , ⁴¹ , ⁴² However, the role of FAAH in osteoclast differentiation is rarely reported. In contrast, two other members of the endocannabinoid system, eCB, CB1 and CB2, have been reported to play an important role in bone homeostasis. Studies have shown that AEA and 2‐AG are produced by osteoblasts and osteoclasts in the skeletal microenvironment. ⁴³ , ⁴⁴ Animal studies suggested that cannabinoids could prevent osteoporosis due to OVX and glucocorticoid injections. ⁴⁵ , ⁴⁶ While another study showed that osteoclast differentiation in humans was associated with increased levels of AEA and decreased levels of 2‐AG. ⁴⁷ In addition, studies have shown the cannabinoid receptors CB1 and CB2 also play a key role in regulating bone homeostasis. Several reports have shown that inhibition of CB1 could inhibit osteoclastogenesis and enhance osteoblast differentiation. ²⁵ , ⁴⁷ , ⁴⁸ , ⁴⁹ However, the role of CB2 in the regulation of bone homeostasis remains controversial. Multiple studies have shown that CB2 plays an important role in promoting the differentiation of both osteoblasts and osteoclasts. ⁵⁰ , ⁵¹ , ⁵² However, other studies have shown that CB2 activation can inhibit osteoclast differentiation while promoting osteogenic differentiation in viro and in vivo. ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶ Based on the role of endocannabinoid ligands and receptors in bone homeostasis, the important role of FAAH as a key enzyme mainly responsible for the hydrolysis of AEA and regulating endocannabinoid signaling transduction, it is reasonable to speculate that FAAH may also be involved in osteoclast differentiation. Therefore, we confirmed that FAAH inhibition can inhibit the differentiation and function of osteoclasts by genetic or pharmacological inhibition of FAAH in BMMs, thus clarifying the positive regulatory role of FAAH in RANKL‐induced osteoclast differentiation.

We further demonstrated the ability of the FAAH inhibitor PF to protect mice from OVX‐induced loss in vivo. μCT results showed that PF partially alleviated OVX‐induced bone loss in mice, and this effect was validated in both the limb bones (distal femur) and axial bone (5th lumbar vertebra), although it was more pronounced in the limb bones. The results of TRAP staining further demonstrated that PF could also significantly inhibit osteoclast differentiation in vivo, which was well consistent with the in vitro experiments. In terms of drug safety, we first proved that PF had no obvious cytotoxicity, and no adverse reactions or deaths were observed in mice during the whole animal experiment. Weekly weight measurement also found that PF administration did not negatively affect the body weight gain of mice. In fact, PF is the first FAAH inhibitor to enter Phase II clinical trials and is considered to be a safe experimental drug with no on‐target toxicity. ⁵⁷ , ⁵⁸ Pharmacokinetic studies in rats, dogs, and humans have shown that PF has good absorption, distribution, homeostasis, and excretion (ADME) properties and is suitable for oral administration. ⁵⁹ After oral administration of PF (0.1 mg/kg), FAAH was almost completely inhibited (>98%) and showed excellent ADME properties in human pharmacokinetics. ⁶⁰

Considering that bone homeostasis is maintained by both osteoblast‐mediated bone formation and osteoclast‐mediated bone resorption, we also preliminarily explored the role of FAAH in osteogenic differentiation. However, the results of ALP staining and ARS staining showed that FAAH inhibitor PF had no significant effect on the differentiation of MC3T3‐E1 cells into osteoblasts. Consistent with this, PF had no significant effect on the expression of osteogenic differentiation marker genes, although the expression of which showed a certain upward trend with the increase of PF concentration. Moreover, the immunohistochemical staining of bone tissue sections and serum ELISA showed that PF had no significant improvement on the decrease of bone formation marker COL1 in bone and P1NP in serum induced by OVX. μCT scan and reconstruction of the cortical bone from the femoral midshaft further confirmed that the inhibition of FAAH had no significant effect on midshaft cortical. These results confirmed that osteogenic differentiation was truly not affected by FAAH inhibition, which was not consistent with the regulation of eCB and receptors CB1 and CB2 involved in osteogenic differentiation. We speculated that the accumulation of AEA caused by FAAH inhibition was not enough to trigger the process of osteogenic differentiation.

Previous studies have confirmed that both CB1 and CB2 play important roles in bone homeostasis. Therefore, to explore the mechanism of FAAH involved in osteoclast differentiation, we first thought that FAAH might act through CB1 or CB2 receptors in the classical endocannabinoid system. Our results showed that neither knockout of FAAH nor its inhibitor had a significant effect on the expression of CB1 and CB2. In fact, there is no report of the interaction between FAAH and receptors CB1 or CB2 in previous studies, and the activation of receptors is not necessarily reflected in the change of expression level. Further rescue experiments proved that neither CB1 antagonist AM251 nor CB2 antagonist AM630 could rescue the inhibitory effect of PF on osteoclast formation and osteoclast marker gene expression. This result was consistent with previous research showing that selective blockade of CB1 (100 nM AM251) and CB2 (100 nM AM630) receptors failed to reverse the inhibitory effect of AEA (10 μM) or 2‐AG (10 μM) on osteoclast differentiation. ⁴⁷ These results suggested that the involvement of FAAH in osteoclast differentiation was independent of CB1 and CB2 receptors activation, and may be mediated by other receptors or a non‐receptor‐mediated effect.

Osteoclast differentiation is regulated by complex signaling pathways, especially the NF‐κB and MAPK pathways. ⁶¹ , ⁶² NF‐κB is a key transcription factor that plays a crucial role in regulating inflammatory responses, and its activation is required for adequate osteoclast differentiation. ⁶³ Previous research has shown that FAAH inhibitors can reduce the oxidative state by reducing the DNA‐binding activity of NF‐κB p65 without cytotoxicity. ⁶⁴ Our results showed that both genetic and pharmacological inhibition of FAAH could inhibit RANKL‐induced NF‐κB and MAPK signaling pathways, suggesting that FAAH inhibition can suppress RANKL‐induced osteoclast differentiation by inhibiting NF‐κB and MAPK pathways activation. It is worth noting that the process of signal transduction is transient, the expression of signaling molecules increased rapidly after stimulation, reached a peak at about 15 min, and then quickly decreased. Therefore, when detecting the changes in signal pathways, time nodes such as 0, 15, 30, and 60 min were selected, which was consistent with the previous studies. ⁶⁵ , ⁶⁶ To further investigate the possible mechanism of FAAH involved in osteoclast differentiation, the whole transcriptome RNAseq was performed. It was encouraging that the IL17 signaling pathway and the osteoclast differentiation pathway were found to be the top two enrichment pathways in KEGG pathway analysis, which was further confirmed by GSEA analysis. Notably, PF intervention significantly down‐regulated the osteoclast differentiation pathway and the osteoclast differentiation marker genes, as well as the classic NF‐κB pathway during osteoclast differentiation, again confirming that FAAH was involved in the process of osteoclast differentiation. Most importantly, we detected that the IL17 signaling pathway was most obviously regulated by PF intervention, even more significantly than the osteoclast differentiation pathway. Multiple genes in the IL17 signaling pathway were also significantly downregulated by PF and further verified by qPCR assay. These results suggested that the IL17 signaling pathway, as a downstream signal of FAAH, mediated the role of FAAH in osteoclast differentiation.

IL17 is a highly versatile pro‐inflammatory cytokine that interacts with the receptors to activate various downstream responses including inflammatory, tissue repair, host defense, and the progression of cancer. ⁶⁷ The IL17 receptor is widely expressed in multiple cell types, including osteoblasts, osteoclasts, fibroblasts, chondrocytes, synoviocytes, monocytes/macrophages, and mast cells. ⁶⁸ , ⁶⁹ , ⁷⁰ IL17 binds to its receptor complex, then recruits Act1 and activates the classical IL17 signaling cascade through TRAF6. TRAF6 binding subsequently triggers the MAPK and NF‐κB pathways. ⁷¹ Notably, TRAF6 is also a key downstream mediator of RANKL in osteoclast differentiation. ⁷² Numerous studies have shown that IL17 is involved in the pathophysiological processes of several diseases, such as breast cancer, lung cancer, inflammatory bowel disease, cardiovascular diseases, uveitis, rheumatoid arthritis, and psoriasis. ⁷³ , ⁷⁴ , ⁷⁵ , ⁷⁶ , ⁷⁷ , ⁷⁸ IL17 also plays an important role in bone homeostasis. Previous studies have shown that IL17 stimulates osteoblasts to produce RANKL, which in turn induces osteoclast progenitor cells to differentiate into mature osteoclasts. ⁷⁹ Local injection of IL17 induced periodontal inflammation, stimulated osteoclast formation and periodontal bone loss in normal germ‐free mice. ⁸⁰ Another study suggested that IL17 could promote osteoclast differentiation by inducing autophagy and aggravate alveolar bone resorption in an experimental periodontitis rat model. ⁸¹ However, the relationship between IL17 signaling and FAAH has not been reported. To further confirm the involvement of IL17 signaling in FAAH‐mediated osteoclastogenesis, we performed rescue experiments. The results showed that recombinant murine IL17 could partially restore the inhibitory effects of PF on osteoclast formation, F‐actin ring formation, and osteoclast bone resorption, and also increased the expression of key transcription factors and osteoclast marker genes. In addition, IL17 partially reversed the inhibition of PF on NF‐κB pathway and MAPK pathway, especially on NF‐κB pathway. Notably, intervening RANKL‐induced BMMs with IL17 alone showed a trend of promoting osteoclast differentiation, but the difference is not statistically significant, which was similar to a previous study, which showed that IL17 had no significant effect on osteoclast formation induced by RANKL, but promoted osteoclastogenesis in a co‐culture system, suggesting that IL17 did not act directly on osteoclast precursor cells. ⁸² In addition, previous studies have shown that IL17 is closely related to OVX‐mediated bone loss. In the OVX‐induced bone loss model, the mRNA levels of RANKL, TNF, and IL17 in the intestines and BM of mice were significantly higher than those in the sham‐operated group. ⁸³ Treatment with an anti‐IL17 antibody or inhibition of IL 17R could prevent OVX‐induced bone loss. ⁸⁴ , ⁸⁵ Our results were consistent with these reports, immunohistochemical results showed that the expression of IL17 increased significantly after OVX, while the effect of PF could notably reduce its expression, which was also confirmed by ELISA. Altogether, these results further suggested that the role of FAAH in promoting osteoclast differentiation was partly mediated by the IL17‐NF‐κB/MAPK signaling cascade.

This study also has some limitations. First, we validated the in vivo effect of FAAH activity using the FAAH inhibitor PF rather than a FAAH gene‐conditioned knockout mice model. Second, the in vivo study lacked the SHAM+PF group to illustrate whether PF had an effect on osteoclasts under normal physiological conditions. Third, there was still a lack of further research on the mechanism of the IL17 signaling pathway participating in the effect of FAAH on osteoclast differentiation. We hypothesized that FAAH, as an important hydrolase in the cannabinoid system, might further affect IL17 signaling by hydrolyzing cannabinoids. As reported in previous studies, AEA treatment significantly reduced IL17 production in mice and humans. ⁸⁶ , ⁸⁷ The plasma levels of CCL2 and IL17 were reduced in non‐naïve cannabis users. ⁸⁸ Fourth, we focused on the effect of FAAH on osteoclasts, and only made a preliminary discussion on the effect of FAAH on osteoblast differentiation, without an in‐depth study of related pathways and mechanisms. Fifth, we used the osteoblast cell line MC3T3‐E1 instead of murine primary osteogenic precursor cells or bone marrow mesenchymal stem cells.

5. CONCLUSIONS

In conclusion, our study suggests that FAAH plays an important role in RANKL‐induced osteoclast formation and bone resorption. Pharmacological inhibition or knockdown of FAAH inhibited RANKL‐induced osteoclast differentiation partially by inhibiting the IL17‐NF‐κB/MAPK signaling cascade in vitro, and also attenuated OVX‐induced bone loss by suppressing osteoclast formation in vivo, suggesting that targeting FAAH is a promising candidate strategy for the treatment of osteoclast‐related diseases, especially osteoporosis.

AUTHOR CONTRIBUTIONS

Meipeng Zhu, Qian Guo, and Feng Li designed the study; Meipeng Zhu performed the cell and animal experiments; Yimin Dong, Zijian Dong, Renpeng Peng, and Liangxi Chen participated in the mice ovariectomy experiment; Yayun Zhang, Shimeng Xu, and Pengju Wang contributed to the statistical analysis. Sibo Wang, Haiyang Liu, Honglei Kang, Kehan Song, and Hongjian Zhao prepared the figures. Meipeng Zhu and Qian Guo wrote the manuscript. Jian Liu and Feng Li revised the final manuscript.

DISCLOSURES

The authors declare no competing conflict of interests.

ETHICS STATEMENT

Ethics approval of animal experiments was obtained from Tongji Hospital Ethics Committee of Huazhong University of Science and Technology.

Supporting information

Table S1

Dataset S2

DATA AVAILABILITY STATEMENT

This study includes no data deposited in external repositories.