Theaflavin-3,3′-digallate represses osteoclastogenesis and prevents wear debris-induced osteolysis via suppression of ERK pathway

Accepted Manuscript Theaflavin-3,3′-digallate represses osteoclastogenesis and prevents wear debris-induced osteolysis via suppression of ERK pathway Xuanyang Hu, Zichuan Ping, Minfeng Gan, Yunxia Tao, Liangliang Wang, Jiawei Shi, Xiexing Wu, Wen Zhang, Huilin Yang, Yaozeng Xu, Zhirong Wang, Dechun Geng PII: DOI: Reference:

S1742-7061(16)30609-2 http://dx.doi.org/10.1016/j.actbio.2016.11.022 ACTBIO 4531

To appear in:

Acta Biomaterialia

Received Date: Revised Date: Accepted Date:

8 August 2016 10 October 2016 8 November 2016

Please cite this article as: Hu, X., Ping, Z., Gan, M., Tao, Y., Wang, L., Shi, J., Wu, X., Zhang, W., Yang, H., Xu, Y., Wang, Z., Geng, D., Theaflavin-3,3′-digallate represses osteoclastogenesis and prevents wear debris-induced osteolysis via suppression of ERK pathway, Acta Biomaterialia (2016), doi: http://dx.doi.org/10.1016/j.actbio. 2016.11.022

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title page Title of the article Theaflavin-3,3′-digallate

represses

osteoclastogenesis

and

prevents

wear

debris-induced osteolysis via suppression of ERK pathway The name(s) of the author(s) Xuanyang Hu a,1 ; Zichuan Pinga,1; Minfeng Gan a,1; Yunxia Taoa,1; Liangliang Wanga; Jiawei Shia; Xiexing Wu a; Wen Zhangb,; Huilin Yanga; Yaozeng Xu

a,*

; Zhirong

Wangc,*; Dechun Genga,* Name of the institution: a: The First Affiliated Hospital of Soochow University. b: Soochow University. c: Zhangjiagang Hospital of Traditional Chinese Medicine *Corresponding authors: Yaozeng Xu (E-mail: [email protected]), Zhirong Wang (E-mail: [email protected]), Dechun Geng (E-mail: [email protected]) 1

Contribute equally to this work

1

Theaflavin-3,3′-digallate

represses

osteoclastogenesis

and

prevents

wear

debris-induced osteolysis via suppression of ERK pathway Abstract Peri-implant osteolysis (PIO) and the following aseptic loosening is the leading cause of implant failure. Emerging evidence suggests that receptor activator of nuclear factor kappa-B ligand (RANKL)-induced osteoclast formation and osteoclastic bone resorption are responsible for particle-stimulated PIO. Here, we explored the effect of theaflavin-3,3′-digallate (TF3) on titanium particle-induced osteolysis in vivo and in vitro. Twenty-eight male C57BL/6 mice were randomly separated into four groups: sham control (sham), titanium particles only (titanium), titanium particles with low TF3 concentration (low-TF3, 1mg/kg TF3), and titanium particles with high TF3 concentration (high-TF3, 10mg/kg TF3). Two weeks later, micro-computed tomography and histological analysis were performed. Bone-marrow-derived macrophages and RAW264.7 murine macrophages were applied to examine osteoclast formation and differentiation. TF3 significantly inhibited titanium particle-induced osteolysis and prevented bone destruction compared with titanium group. Interestingly, the number of mature osteoclasts reduced after treatment with TF3 in vivo, suggesting osteoclast formation might be inhibited by TF3. In vitro, TF3 suppressed osteoclast formation, polarization and osteoclastic bone resorption by specifically targeting the RANKL-induced ERK signal pathway. Collectively, these results suggest that TF3, a natural active compound derived from black tea, is a promising candidate for the treatment of osteoclast-related osteolytic diseases, such as wear debris-induced PIO. Keywords: peri-implant osteolysis; wear debris; theaflavin-3,3′-digallate; RANKL; ERK pathway 2

1.

Introduction Total joint arthroplasty (TJA) is widely accepted for the treatment of end-stage

joint diseases such as severe trauma, osteoarthritis and rheumatoid arthritis[1]. However, it is reported that aseptic loosening, initiated by peri-implant osteolysis (PIO), is the major reason for prosthesis failure[2, 3], resulting in more than one-third of patients undergoing revision surgery within two decades after TJA[4]. Although revision surgery has a positive effect on aseptic loosening, we cannot overlook the negative impacts of revision surgery, such as higher cost, shorter survival duration and poorer clinical outcomes. Although the pathophysiology of PIO remains unclear, increasing evidence indicates that osteoclasts are excessively activated at the implant site by wear debris from materials such as titanium (Ti), polymethyl methacrylate, polyethylene, and cobalt chromium[5, 6]. Two essential factors are reported for osteoclastogenesis: macrophage colony-stimulating factor (M-CSF) and receptor activator of nuclear factor kappa-B (RANK) ligand (RANKL). M-CSF is an important regulator for the proliferation and survival of osteoclast precursor cells and up-regulates the expression of RANK, which is vital for the differentiation of osteoclast precursor cells into mature osteoclasts[7-9]. Binding of RANKL to its receptor, RANK, promptly activates a number of signaling pathways such as mitogen-activated protein kinases (MAPKs), nuclear factor-κB (NF-κB) and phosphatidylinositol 3-kinase/AKT (PI3K/Akt)[10-13]. This in turn leads to the activation of downstream signaling factors including activated protein 1 (AP1) and nuclear factor of activated T-cells 1 (NFATc1), which are crucial for the formation of osteoclasts and expression of osteoclast-related genes[14]. Considering the critical role of osteoclasts in osteolytic diseases, inhibitors that specifically target the suppression of osteoclast formation and/or function are ideal 3

candidates to prevent wear debris-induced PIO and bone loss. In recent years, attention has been paid to more natural compounds with the potential to inhibit osteoclastogenesis and osteoclast-related diseases. Our research group has screened a number of natural compounds to identify candidates to treat osteoclast-related osteolytic diseases, including wear debris-induced PIO. Black tea, extracted from the leaves of Camellia sinensis, is one of the most popular beverages worldwide[15]. The fermentation process, which is indispensable in the process of making black tea, results in oxidation of polyphenols and subsequent formation of oligomeric flavanols containing theaflavins, thearubigin and other oligomers[16]. Compared with green tea, which is rich in catechin, black tea has a high content of tea theaflavins. Of the four main theaflavins found in black tea, theaflavin-3,3′-digallate (TF3) is formed by polyreaction of epicatechin gallate (ECG) and (−)-epigallocatechin-3-gallate (EGCG), which are the major catechins found in green tea[17]. Previous studies have reported that similar to EGCG, TF3 has many positive effects on human health, including antioxidant[18], anti-angiogenic[17] and anti-tumor properties[19]. Recently, TF3 has been reported to inhibit osteoclast formation[20] and prevent ovariectomy-induced bone loss[21]. However, whether TF3 can prevent wear debris-induced PIO remains unknown. Because osteoclasts play a crucial role in PIO[5, 6], it is logical that TF3 could be an optimal treatment candidate. We hypothesized that TF3 acts to prevent debris-induced PIO through the inhibition of osteoclast formation to prevent bone loss. The aims of this study were therefore to investigate the effect of TF3 on particle-induced PIO in a mouse calvarial model and osteoclasts formation in vitro, as well as the underlying mechanism of TF3 during osteoclastogenesis. 2. Materials and Methods 4

2.1.Preparation of metal particles Commercially pure titanium (Ti) particles (catalog #00681) were obtained from Johnson Matthey chemicals (Ward Hill, Massachusetts, USA). A Coulter counter™ (BeckmanCoulter Inc., USA) with five consecutive measurements was used to test the distribution and size of the Ti particles. Ninty precents of the Ti particles were <3.6 mm. As previously described[22, 23], it was used to eliminate endotoxins that the particles were sterilized at 180 ℃ for 6 h and washed in 75% ethanol for 48h. Endotoxin levels were confirmed by a quantitative Limulus Amebocyte Lysate Assay (Bio-Whittaker, Walkersville, MD, USA) and only free of endotoxin particles were used. 2.2. Experimental animals and drug treatment All experiments were performed in strict accordance with the guidelines for Care and Use of Laboratory Animals and were approved by the Animal Care Committee of our institute. Based on the previous studies [24-26], we established a murine calvarial osteolysis model with Ti particle-stimulation. Briefly, 28 6- to 8-week-old male C57BL/6 mice were separated at random into four groups (n=7): (1) sham group (underwent sham surgery only), (2) Ti group (received 30mg Ti particles to induce osteolysis), (3) low-TF3 group (received 30mg Ti particles with low TF3 concentration, Sigma-Aldrich, St. Louis, MO, USA), and (4) high-TF3 group (received 30mg Ti particles with high TF3 concentration). According to the products instruction, TF3 was resuspended in absolute alcohol until fully dissolved and then diluted in phosphate-buffered saline (PBS). Mice in the low- and high-TF3 groups were intraperitoneally injected with TF3 (1mg/kg and 10mg/kg, respectively) every second day for 2 weeks, wheras the rest of groups were injected with PBS. For endotoxins elimination, Ti particles were prepared according to the method established by Geng 5

et al [24]. The animals were sacrificed after 2 weeks of Ti particle implantation and the calvariae were removed, fixed and prepared for further analysis. 2.3. µCT scanning After fixed with 10% formaldehyde for 24h, the mice calvariae were analyzed with a SkyScan1176 µCT. The calvariae (n=4) were scanned at an equidistant resolution of 18µm with 80kV and 100µA energy X-ray. To avoid metal artifacts, wear particles were removed before scanning. Images were reconstructed using the software provided by the manufacturer. As previously described[27], the combined region of interest (ROI; 3×3×1mm) with the midline suture at its center was identified for further quantitative analysis. Bone mineral density (BMD, mg/cc), number and area of pores in each sample and the ratio of body volume to tissue volume (BV/TV, %) were measured by CT Analyser software (Skyscan). 2.4. Histological and histomorphometric analysis. The calvariae(n=3, per group) were fixed in 4% paraformaldehyde (PFA) for 48h, followed by decalcification in 10% EDTA for 3 weeks and paraffin embedding. Sections (5µm) were cut and prepared for hematoxylin and eosin (H&E) and tartrate-resistant acid phosphatase (TRAP) staining as previously described [26]. The TRAP staining was performed using a commercial TRAP kit (#387A, Sigma-Aldrich). Photos of the stained sections were observed in a Zeiss microscope (Zeiss, Dreseden, Germany). The eroded surface (mm2), periosteum thickness (mm), osteoclasts number and the ratio of osteoclast to bone surface (OCs/BS, %) were measured and quantified with the help of image analysis software (Image Pro-Plus 6.0) according to the methods established by Kauther [28]. 2.5. Cell extraction and osteoclast differentiation

6

To prepare osteoclast precursors, bone-marrow-macrophages (BMMs) were extracted from the femurs and tibiae of 4- to 6-week-old C57/BL6 mice. Briefly, cells were separated and cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% FBS, 100 U/ml penicillin and M-CSF (30ng/ml) (R&D Systems, Minneapolis, MN, USA) for 16h. The cells adherent to the bottom of cell culture plates were removed and the non-adherent cells were incubated in a 37°C/5% CO2 incubator in 6-well plates for another 3 days until fully confluent. Subsequently, we removed the suspension cells and the cells adherent to the bottom of cell culture plates were used as BMMs. The BMMs (1×104 cells/well) were added to 96-well plates as previously described. After treatment with or without different concentrations of TF3 (0, 0.01, 0.1, 1µM) for 4h, the cells were cultured in induction medium containing 30ng/mL M-CSF and 50ng/mL RANKL (R&D Systems) for 5 days. The medium was changed every 3 days. At the specified time points, the medium was removed. After washed three times with PBS for 3min, fixed with 4% PFA for 10min, the cells were dyed for TRAP according to the instruction of TRAP Kit. The photos of TRAP-positive cells with nuclei ≧ 3 were obtained using a high-quality light microscope (Zeiss). 2.6. Cell viability assay To detect the toxic effect of TF3 on BMMs, the cell counting kit-8(CCK-8, Dojindo,Kumamoto, Japan) assay was used. After plated on 96-well plates in triplicate, BMMs (1.5×104 cells/well) were cultured in induction medium. After 24h, the medium was removed and various concentrations of TF3 (0.01-40µM) were added for 48h. Then, 100µl of fresh medium containing 10µl CCK-8 were added into the well. After incubation at 37°C for 2h, cell viability was determined using a microplate reader(BioTek, Vermont, USA) to read the optical density at an absorbance of 450nm. 7

2.7. Resorption pit assay To examine osteoclast function, BMMs were seeded into a 24-well bone resorption plate (OAP, Corning, New York, USA) at a density of 1×105/well in triplicate. After adhering, cells were pre-treated for 4h with various concentrations of TF3 in basal medium supplemented with 30ng/ml M-CSF. The BMMs were then induced into osteoclasts in induction medium for 5 days. The plates were subsequently washed by sonication until cells were completely disrupted and removed from the bottle of plates. The visual images of resorption pits were snapped at the magnification of 5× using an inverted microscope (Zeiss) and Image J software was used to quantify the area of bone resorption. 2.8. F-actin ring formation assay To carry out fluorescence staining of the F-actin ring, osteoclasts treated with or without various concentration of TF3 were fixed in 4% PFA for 10min, permeabilized in 0.1% Triton X-100/PBS for 5min, and then incubated with Acti-stain 488 Fluorescent Phalloidin (3.5µl solution+500µl PBS, Cytoskeleton Inc, L.A ., USA) at room temperature in the dark for 30min. After being washed three times with PBS, the cell nucleus was counterstained with 4',6-diamidino-2-phenylindole (DAPI, Solarbio, Beijing, China) for 10min. Osteoclastic rings were visualized using an immunofluorescence microscope (Zeiss). 2.9. Quantitative real-time PCR analysis Total RNA was extracted using TRIzol (Sigma-Aldrich) and RNA was reversely transcribed to cDNA with reverse transcriptase (Takara, Otsu, Japan). For real-time PCR, cDNAs were augmented with the SYBR Premix Ex Tag Kit (Takara). All PCR samples were run in triplicate and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control. All the mouse primer sequences for TRAP, 8

NFATc1, c-Fos, cathepsin K (CTSK), osteoclast associated receptor (Oscar) and GAPDH

were

as

follows:

TRAP

5’-CTGGAGTGCACGATGCCAGCGACA-3’

forward

and

reverse

5’-

TCCGTGCTCGGCGATGGACCAGA-3’;

NFATc1

forward

5’-

CCGTTGCTTCCAGAAAATAACA-3’

and

reverse

5’-

TGTGGGATGTGAACTCGGAA-3’;

c-Fos

forward

5’-

CCAGTCAAGAGCATCAGCAA-3’

and

reverse

5’-

AAGTAGTGCAGCCCGGAGTA3’;

CTSK

forward

5’-

CTTCCAATACGTGCAGCAGA-3’

and

reverse

5’-

TCTTCAGGGCTTTCTCGTTC-3’;

Oscar

forward

5’-

CTGCTGGTAACGGATCAGCTCCCCAGA-3’

and

CCAAGGAGCCAGAACCTTCGAAACT-3’; ACCCAGAAGACTGTGGATGG-3’

GAPDH and

reverse

5’-

forward

5’-

reverse

5’-

CACATTGGGGGTAGGAACAC-3’. 2.10. Western blot analysis Western blotting was performed as described previously[29]. RAW264.7 (5×10 5 cells/well) was plated on 6-well plates. On reaching confluence, cells were pre-treated with or without 1µM TF3 for 4h and then stimulated with 50ng/ml RANKL for 0, 5, 15, 30 or 60min both in control (untreated with TF3) and TF3-treated groups. The cells were then collected and lysed using 50µl radioimmunoprecipitation assay lysis buffer with proteinase and phosphatase inhibitors (Sigma-Aldrich). After lysis, the samples were centrifuged at 12,000×g for 5min and the supernatants were collected to determine the concentration of protein using a bicinchoninic acid protein kit (BCA kit, Sigma-Aldrich). Proteins (30µg) were isolated by 12% SDS-PAGE and transferred to a polyvinylidene fluoride membrane, which was blocked by 5% bovine serum 9

albumin in Tris-buffered saline-Tween (10mM Tris-HCl, 50mM NaCl, 0.25% Tween 20) for 2h and then incubated with primary antibody including β-Actin(1:1000), IκBα(1:1000),

P-IκBα(1:1000),

p65(1:1000),

P-p65(1:1000),

MEK(1:1000),

P-MEK(1:1000), ERK(1:1000), P-ERK(1:2000), JNK(1:1000), P-JNK(1:1000), p38(1:1000),

P-p38(1:1000)

(Cell Signaling

Technology,

MA,

USA)

and

AKT1(1:1000), P-AKT1(1:5000), NFATc1(1:2000), c-Fos(1:5000) (Abcam, Camb, UK) at 4°C overnight. After being washed three times with TBS-Tween, the membranes were incubated with horseradish peroxidase-conjugated secondary antibody.

Proteins

then

appeared

as

fluorescent

bands

using

enhanced

chemiluminescence (ECL) reagent (Sigma-Aldrich). 2.11. Statistical analysis Data are shown as means ± standard deviation (SD). All experiments were performed at least three times independently. One-way analysis of variance (ANOVA) with Tukey post-hoc multiple comparison tests and paired t tests were used for statistical analyses. A P value of <0.05 represented statistical significance. All statistics were calculated using SPSS 17.0 software (SPSS, Chicago, IL, USA). 3. Results 3.1. TF3 exerted a therapeutic effect on Ti particle-induced osteolysis and prevented bone loss in vivo Three-dimensional reconstruction of µCT images showed that calvariae from the Ti particle-stimulated mice were extensively eroded (Fig. 1A). In addition, BMD and BV/TV were significantly reduced (Fig. 1B, C), whereas the number and area of porosity were increased when compared with the sham group (P< 0.01) (Fig. 1D, E). However, after treatment with TF3, Ti particles-stimulated osteolysis in the mouse calvariae were markedly ameliorated and the BMD, BV/TV and porosity values were 10

dramatically reversed in both low- and high-TF3 treated groups (Fig. 1). Consistent with the µCT results, histological staining and histomorphometric analysis further confirmed a protective effect of TF3 on particle-induced bone destruction in vivo (Fig. 2). H&E staining revealed that a large number of lymphocytes and macrophages infiltrated into the calvarial surface in the Ti particle-stimulated mice. Similarly, a lot of TRAP-positive cells were present along the area of eroded bone surface in mice stimulated with Ti particles, as clearly shown by TRAP staining. In contrast, treatment with TF3 considerably reduced the area of bone erosion with osteoclast cell number, and the OcS/BS decreased in a dose-dependent manner. 3.2. TF3 inhibited RANKL-induced osteoclastogenesis in a dose- and time-dependent manner in vitro To determine TF3 toxicity on osteoclast precursor cells (BMMs), the CCK-8 assay was performed. The results showed that a TF3 dose <5µM had no cytotoxic effect on BMMs (Fig. 3). We therefore treated BMMs with induction medium containing 30ng/ml CSF and 50ng/ml RANKL, together with or without the different concentration of TF3 (0.01, 0.1 and 1µM) for 5 days. TRAP staining indicated that mature osteoclasts number decreased in the presence of TF3 from 267.7 ± 32.2 (0µM) per well to 66.6 ± 5.4 (1 µM) per well and the osteoclast area reduced from 74.9% (0µM) to 16.4% (1µM). To examine which stages of osteoclastogenesis were affected, we treated BMMs with 1µM TF3 at 0, 2 or 4-day together with 30ng/ml M-CSF and 50ng/ml RANKL for 5 days. TRAP staining suggested that compared with the control group, osteoclast number and area were markedly reduced by TF3 at 0 and 2 days (Fig. 3), with no significance difference at 4 days. We can therefore conclude that TF3 suppressed the number and size of osteoclasts in the early stages (at 0 and 2 days) rather than later stages (at 4 days). 11

3.3. Osteoclast bone resorptive function and F-actin ring formation were impaired by TF3 in vitro To determine whether TF3 could inhibit osteoclastic bone resorption, BMMs were seeded in 24-well bone plates in the presence or absence of 0.01, 0.1, and 1µM of TF3. The results showed that the area of bone resorption reached 81.9% in the control group. However, after treatment with 1 µM TF3, the bone resorption area was decreased to 25.9% (Fig. 4A, B). For successful osteoclastic bone resorption, it is essential that the F-actin rings are well-polarized[30]. We therefore tested whether TF3 could inhibit formation of the F-actin ring. As expected, staining with Acti-stain 488 Fluorescent Phalloidin suggested that the morphology and size of the F-actin rings were dramatically decreased when treated with TF3 in a dose-dependent manner (Fig. 4C). 3.4. TF3 suppressed the expression of RANKL-induced osteoclast-related genes in vitro Upregulation of osteoclast-related genes plays an important role in osteoclast differentiation. To further assess the effect of TF3 on osteoclastogenesis, we determined

whether

TF3

inhibited

the

expression

of

RANKL-induced

osteoclast-related genes, including NFATc1, c-Fos, TRAP, CTSK, and Oscar. As expected, quantitative real-time PCR showed that after stimulation with 50ng/ml RANKL, the mRNA expression levels of these genes were elevated. However, the effect of RANKL on inducing osteoclast-related genes was strongly inhibited by TF3 in a dose- and time-dependent manner (Fig. 5A, B). Consistent with the TRAP staining results, TF3 demonstrated an inhibitory effect on RANKL-induced osteoclast-related gene expression in vitro. 3.5. TF3 inhibited ERK signaling pathway activation during osteoclastogenesis 12

Several signaling pathways are reported to be associated with the formation of osteoclasts, including MAPKs (ERK1/2, JNK, p38), NF-κB (IkBα, p65) and PI3k/Akt[10-13]. To explore which of these signaling pathways are affected by TF3, western blot analysis was performed. After pre-treatment with TF3 for 4h, RAW264.7 cells were stimulated with 50ng/ml RANKL for the indicated time (0, 5, 15, 30, or 60min). The results showed that TF3 had no inhibitory effect on NF-κB and PI3k/Akt signaling pathways (Fig. 6A, B). Similarly, TF3 had little effect on activation of the MAPK subfamilies, JNK and p38. In contrast, TF3 was found to affect the suppression of ERK activation. In the control group, ERK phosphorylation was activated after stimulation by RANKL for 15min. However, this trend was dramatically inhibited in the TF3-treated groups (Fig. 6C, D). To further confirm that TF3 exerted a pharmacological inhibitory effect on the ERK signaling pathway, we treated RAW264.7 cells with various concentrations of TF3 (0, 0.01, 0.1, or 1µM) for 15min. The results showed that when stimulated with 50ng/ml RANKL alone, phosphorylation of ERK was enhanced. However, after treatment with TF3, this trend was diminished in a dose-dependent manner (Fig. 7A). Subsequently, we investigated whether TF3 could suppress activation of the upstream ERK signaling regulator, MAP kinase 1/2 (MEK1/2), in RAW264.7 cells treated identically to those described above. As shown in Figure 7B, phosphorylation of MEK1/2 was not inhibited by TF3, which suggested that the initial signal target for TF3 during osteoclastogenesis is the ERK rather than MEK1/2 signaling pathway. We next treated BMMs with RANKL in the absence or presence of TF3 for 0, 1, and 3 days to examine the inhibitory effect of TF3 on the expression of c-Fos and NFATc1. Analogous with the results of RT-PCR, western blot analysis indicated that expression

13

of c-Fos and NFATc1 induced by RANKL was obviously attenuated by TF3 (Fig. 7C). 4. Discussion Total joint arthroplasty is an effective procedure for the treatment of end-stage joint diseases[1]. However, aseptic loosening secondary to PIO is a major reason for arthroplasty failure[2,3]. Although implant designs and materials are greatly improved, many approaches fail to eliminate wear particles generated from implant surfaces[31-33]. Although the precise mechanism remains unknown, many research groups have reported that osteoclasts play an important role in the development of PIO[5,6]. Consequently, osteoclasts are considered to be a major target for treating osteolytic diseases. In fact, many agents known to target osteoclast function, such as bisphosphonates[34, 35], strontium ranelate[36] and erythromycin[37], have been demonstrated to suppress wear debris-stimulated bone resorption. However, the adverse reactions of these agents, namely fever, stomach ulcers[38], osteonecrosis of the jaw, and atypical fractures[35, 39, 40], as well as limitations in their bioavailability, prevent their long-term use for the treatment of PIO. In light of this, Traditional Chinese Medicine (TCM) has prompted many investigators to explore alternative compounds with fewer side effects for their effect on osteolytic disease. Our team and other researchers have reported several TCM compounds that may effectively treat particle-induced PIO[29, 41], which has inspired our group to screen other TCM compounds and explore their therapeutic potential. TF3 is the predominant theaflavin in black tea, derived from the leaves of camellia sinensis, and is reported to demonstrate pharmacological activities in human diseases, including antioxidant[18], anti-angiogenic factor[17] and anti-tumor factor[19]. More recently, TF3 has demonstrated beneficial effects on bone 14

metabolism[20, 21]. However, it still remains unknown whether TF3 can represent a therapeutic effect on osteolytic disease-in particular wear debris-induced PIO. Therefore, the present study was designed to determine if TF3 can inhibit wear debris-induced bone loss in a murine calvarial model, and to explore the underlying mechanism of TF3 on osteoclastogensis and osteoclastic bone resorption. As expected, TF3 significantly alleviated the severity of bone destruction in wear debris-stimulated animals. The presence of Ti particles stimulated a distinct bone resorption reaction in the calvariae of mice, in line with previous studies[29]. In contrast, TF3 treatment dose-dependently suppressed Ti particle-induced bone resorption, as illustrated by an increase in BMD and BV/TV, and a decrease in porosity in mouse calvariae. Meanwhile, bone histomorphometry results further demonstrated that TF3 could inhibit Ti particle-induced osteolysis in vivo, supporting our original hypothesis. Taken together, the evidence indicates that TF3 can exert a therapeutic effect to treat wear debris-induced osteolysis. Osteoclasts are known to be critical for the initiation of wear-debris-stimulated bone resorption[5, 6]. We therefore considered whether TF3 exert an inhibitory effect on osteoclast formation. In the current study, TF3 was clearly proven to be an effective treatment for suppressing the number and area of TRAP-positive cells in vitro. In addition, we further confirmed that TF3 suppressed the mRNA levels of osteoclast-related genes, such as NFATc1, c-Fos, TRAP, CTSK, and Oscar. Furthermore, in vivo TF3 treatment obviously decreased the TRAP-positive cells number in a dose-dependent manner when compared to the Ti group. The results in present study are consistent with previous studies in that the threapeutic effects of TF3 and theaflavin-related drugs on osteolysitc disease are related with formation and differentiation of osteoclast [20, 21]. Our data indicate that TF3 exerts its protective 15

effect on particle-stimulated osteolysis through inhibition of osteoclast formation and differentiation. We next investigate the underlying mechanisms of TF3-induced inhibition on osteoclast formation and osteoclastic bone resorption. Our results demonstrated that TF3 impaired the RANKL-induced ERK signal pathway, but that JNK, p38, NF-κB, and PI3k/Akt were not affected. This observation was further supported by TF3 down-regulating the activation of ERK in a dose-dependent manner. Previous studies have revealed the critical role of the ERK signal pathway in initiating wear debris-stimulated osteolysis and osteoclast formation[9, 11, 42, 43]. RANKL is a powerful factor of activating the MAPK superfamily, containing ERK, which is phosphorylated preferentially by MEK1/2. Phosphorylated ERK subsequently promotes the activation c-Fos. This factor is a component of AP-1, which is an essential translation factor during osteoclast formation. In addition, c-Fos is vital for NFATc1 induction and translocation. Previous studies have demonstrated that NFATc1 is a master transcription factor that is critical for osteoclast formation and activation[29, 43]. Interestingly, we demonstrated that TF3 significantly reduced the protein and mRNA levels of c-Fos and NFATc1, even in the presence of RANKL, which subsequently inhibited the expression of osteoclast-related genes including TRAP, CTSK and Oscar in a dose- and time-dependent manner. In an additional study, MEK1/2 was unaffected by TF3 treatment, even though ERK phosphorylation was dramatically inhibited. This indicated that TF3 acts to inhibit activation of the ERK signaling pathway without affecting upstream regulators. Taken together, these results suggest TF3 acts to suppress activation of the ERK/c-Fos/NFATc1 signaling pathway during osteoclastogenesis, with no obvious effect on p38, JNK, NF-κB and PI3k/Akt activation. 16

Despite these interesting findings, our study has some limitations. First, bone metabolism is a balance between osteoclastic bone resorption and osteoblastic bone formation[44, 45]. Bone formation also plays an important role in treating osteolysis[46]. To date, we have not explored whether TF3 can enhance osteoblastogenesis, and thus improve bone formation. This will be investigated in future studies. Second, the main cause of clinical PIO is polyethylene rather than Ti particles[47]. The reason we choose to use Ti particles is that they are well-characterized. However, it has been reported that both Ti and polyethylene comparably imitate authentic wear particles to induce osteolysis[48, 49]. Finally, we used a Ti particle-induced mouse calvarial model to generate osteolysis. Von Knoch et al reported that particles in patients are continuously released from the surface of the implant, differing from the single particle exposure experienced in the mouse osteolysis model[49]. In addition, calvarial and long bone structures differ with regard to cell components, bone micro-architecture and mineralization. Therefore, a rat femur titanium rod model may be more adaptable for mimicking osteolysis. This model will be assessed in future studies. 5. Conclusions In summary, the results of the present study demonstrate that TF3 inhibited osteoclastogenesis and the expression of osteoclast-related genes mainly via suppressing the ERK signaling pathway and its downstream factors including NFATc1 and c-Fos. Moreover, the findings of this study have confirmed for the first time that TF3 has a protective effect on Ti particle-induced osteolysis in a mouse calvarial model, thus preventing bone loss. These results indicate that TF3 might be a suitable therapeutic agent for the treatment of wear-debris-induced PIO. Acknowledgments 17

We thank Prof. An Qin for technical help and advice. This research was supported by the National High Technology Research and Development Program of China (2015AA020316), the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), the Nature Science Foundation of Jiangsu Province (BK2011303) and the National Nature Science Foundation of China (81101399, 81272018, 81372018, 81472077 and 81672238). References [1] Urban RM, Hall DJ, Della Valle C, Wimmer MA, Jacobs JJ, Galante JO. Successful long-term fixation and progression of osteolysis associated with first-generation cementless acetabular components retrieved post mortem. J Bone Joint Surg Am. 2012;94:1877-85. [2] Wooley PH, Schwarz EM. Aseptic loosening. Gene Ther. 2004;11:402-7. [3] Holt G, Murnaghan C, Reilly J, Meek RM. The biology of aseptic osteolysis. Clin Orthop Relat Res. 2007;460:240-52. [4] Keener JD, Callaghan JJ, Goetz DD, Pederson DR, Sullivan PM, Johnston RC. Twenty-five-year results after Charnley total hip arthroplasty in patients less than fifty years old: a concise follow-up of a previous report. J Bone Joint Surg Am. 2003;85-a:1066-72. [5] Masui T, Sakano S, Hasegawa Y, Warashina H, Ishiguro N. Expression of inflammatory cytokines, RANKL and OPG induced by titanium, cobalt-chromium and polyethylene particles. Biomaterials. 2005;26:1695-702. [6] Purdue PE, Koulouvaris P, Potter HG, Nestor BJ, Sculco TP. The cellular and molecular

biology

of

periprosthetic

osteolysis.

2007;454:251-61.

18

Clin

Orthop

Relat

Res.

[7] Koga T, Inui M, Inoue K, Kim S, Suematsu A, Kobayashi E, et al. Costimulatory signals mediated by the ITAM motif cooperate with RANKL for bone homeostasis. Nature. 2004;428:758-63. [8] Geng DC, Xu YZ, Yang HL, Zhu GM, Wang XB, Zhu XS. Cannabinoid receptor-2 selective antagonist negatively regulates receptor activator of nuclear factor kappa B ligand mediated osteoclastogenesis. Chin Med J (Engl). 2011;124:586-90. [9] Suda T, Kobayashi K, Jimi E, Udagawa N, Takahashi N. The molecular basis of osteoclast differentiation and activation. Novartis Found Symp. 2001;232:235-47; discussion 47-50. [10] Takayanagi H. Osteoimmunology: shared mechanisms and crosstalk between the immune and bone systems. Nat Rev Immunol. 2007;7:292-304. [11] Leibbrandt A, Penninger JM. RANK/RANKL: regulators of immune responses and bone physiology. Ann N Y Acad Sci. 2008;1143:123-50. [12] Mandal CC, Ghosh Choudhury G, Ghosh-Choudhury N. Phosphatidylinositol 3 kinase/Akt

signal

relay

cooperates

with

smad

in

bone

morphogenetic

protein-2-induced colony stimulating factor-1 (CSF-1) expression and osteoclast differentiation. Endocrinology. 2009;150:4989-98. [13] Park-Min KH, Lim E, Lee MJ, Park SH, Giannopoulou E, Yarilina A, et al. Inhibition of osteoclastogenesis and inflammatory bone resorption by targeting BET proteins and epigenetic regulation. Nat Commun. 2014;5:5418. [14] Asagiri M, Takayanagi H. The molecular understanding of osteoclast differentiation. Bone. 2007;40:251-64. [15] Yang CS, Landau JM. Effects of tea consumption on nutrition and health. J Nutr. 2000;130:2409-12. 19

[16] Liang YC, Chen YC, Lin YL, Lin-Shiau SY, Ho CT, Lin JK. Suppression of extracellular signals and cell proliferation by the black tea polyphenol, theaflavin-3,3'-digallate. Carcinogenesis. 1999;20:733-6. [17] Gao Y, Rankin GO, Tu Y, Chen YC. Theaflavin-3, 3'-digallate decreases human ovarian carcinoma OVCAR-3 cell-induced angiogenesis via Akt and Notch-1 pathways, not via MAPK pathways. Int J Oncol. 2016;48:281-92. [18] Leung LK, Su Y, Chen R, Zhang Z, Huang Y, Chen ZY. Theaflavins in black tea and

catechins

in

green tea

are

equally

effective

antioxidants.

J

Nutr.

2001;131:2248-51. [19] Kundu T, Dey S, Roy M, Siddiqi M, Bhattacharya RK. Induction of apoptosis in human leukemia cells by black tea and its polyphenol theaflavin. Cancer Lett. 2005;230:111-21. [20] Oka Y, Iwai S, Amano H, Irie Y, Yatomi K, Ryu K, et al. Tea polyphenols inhibit rat osteoclast formation and differentiation. Journal of Pharmacological Sciences. 2012;118:55-64. [21] Nishikawa K, Iwamoto Y, Kobayashi Y, Katsuoka F, Kawaguchi S, Tsujita T, et al. DNA methyltransferase 3a regulates osteoclast differentiation by coupling to an S-adenosylmethionine-producing metabolic pathway. Nat Med. 2015;21:281-7. [22] Ragab AA, Van De Motter R, Lavish SA, Goldberg VM, Ninomiya JT, Carlin CR, et al. Measurement and removal of adherent endotoxin from titanium particles and implant surfaces. J Orthop Res. 1999;17:803-9. [23] Geng D, Xu Y, Yang H, Wang J, Zhu X, Zhu G, et al. Protection against titanium particle induced osteolysis by cannabinoid receptor 2 selective antagonist. Biomaterials. 2010;31:1996-2000.

20

[24] Lu YC, Chang TK, Yeh ST, Fang HW, Lin CY, Hsu LI, et al. The potential role of strontium ranelate in treating particle-induced osteolysis. Acta Biomater. 2015;20:147-54. [25] Liu X, Qu X, Wu C, Zhai Z, Tian B, Li H, et al. The effect of enoxacin on osteoclastogenesis and reduction of titanium particle-induced osteolysis via suppression of JNK signaling pathway. Biomaterials. 2014;35:5721-30. [26] Geng D, Wu J, Shao H, Zhu S, Wang Y, Zhang W, et al. Pharmaceutical inhibition of glycogen synthetase kinase 3 beta suppresses wear debris-induced osteolysis. Biomaterials. 2015;69:12-21. [27] Yang H, Xu Y, Zhu M, Gu Y, Zhang W, Shao H, et al. Inhibition of titanium-particle-induced inflammatory osteolysis after local administration of dopamine and suppression of osteoclastogenesis via D2-like receptor signaling pathway. Biomaterials. 2016;80:1-10. [28] Kauther MD, Bachmann HS, Neuerburg L, Broecker-Preuss M, Hilken G, Grabellus F, et al. Calcitonin substitution in calcitonin deficiency reduces particle-induced osteolysis. BMC Musculoskeletal Disorders. 2011;12:186. [29] Ouyang Z, Zhai Z, Li H, Liu X, Qu X, Li X, et al. Hypericin suppresses osteoclast formation and wear particle-induced osteolysis via modulating ERK signalling pathway. Biochem Pharmacol. 2014;90:276-87. [30] Wilson SR, Peters C, Saftig P, Bromme D. Cathepsin K activity-dependent regulation of osteoclast actin ring formation and bone resorption. J Biol Chem. 2009;284:2584-92. [31] Mutimer J, Devane PA, Adams K, Horne JG. Highly crosslinked polyethylene reduces wear in total hip arthroplasty at 5 years. Clin Orthop Relat Res. 2010;468:3228-33. 21

[32] Muratoglu OK, O'Connor DO, Bragdon CR, Delaney J, Jasty M, Harris WH, et al. Gradient crosslinking of UHMWPE using irradiation in molten state for total joint arthroplasty. Biomaterials. 2002;23:717-24. [33] Dorr LD, Wan Z, Shahrdar C, Sirianni L, Boutary M, Yun A. Clinical performance of a Durasul highly cross-linked polyethylene acetabular liner for total hip arthroplasty at five years. J Bone Joint Surg Am. 2005;87:1816-21. [34] Weinstein RS, Roberson PK, Manolagas SC. Giant osteoclast formation and long-term oral bisphosphonate therapy. N Engl J Med. 2009;360:53-62. [35] Verron E, Bouler JM. Is bisphosphonate therapy compromised by the emergence of adverse bone disorders? Drug Discov Today. 2014;19:312-9. [36] Reginster JY, Lecart MP, Deroisy R, Lousberg C. Strontium ranelate: a new paradigm in the treatment of osteoporosis. Expert Opin Investig Drugs. 2004;13:857-64. [37] Gannon SC, Cantley MD, Haynes DR, Hirsch R, Bartold PM. Azithromycin suppresses human osteoclast formation and activity in vitro. J Cell Physiol. 2013;228:1098-107. [38] Qin XY, Pilot MA, Thompson HH, Maskell JP. Comparison of the side effects and gastrointestinal motility observed after administration of erythromycin and josamycin to dogs. J Antimicrob Chemother. 1986;18:747-56. [39] Yamashita J, McCauley LK. Antiresorptives and osteonecrosis of the jaw. J Evid Based Dent Pract. 2012;12:233-47. [40] Vives V, Cres G, Richard C, Busson M, Ferrandez Y, Planson AG, et al. Pharmacological inhibition of Dock5 prevents osteolysis by affecting osteoclast podosome organization while preserving bone formation. Nat Commun. 2015;6:6218.

22

[41] Yuan FL, Xu RS, Jiang DL, He XL, Su Q, Jin C, et al. Leonurine hydrochloride inhibits osteoclastogenesis and prevents osteoporosis associated with estrogen deficiency by inhibiting the NF-kappaB and PI3K/Akt signaling pathways. Bone. 2015;75:128-37. [42] Monje P, Hernandez-Losa J, Lyons RJ, Castellone MD, Gutkind JS. Regulation of the transcriptional activity of c-Fos by ERK. A novel role for the prolyl isomerase PIN1. J Biol Chem. 2005;280:35081-4. [43] Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berman K, et al. Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr Rev. 2001;22:153-83. [44] Sims NA, Martin TJ. Coupling the activities of bone formation and resorption: a multitude of signals within the basic multicellular unit. Bonekey Rep. 2014;3:481. [45] Li D, Liu J, Guo B, Liang C, Dang L, Lu C, et al. Osteoclast-derived exosomal miR-214-3p inhibits osteoblastic bone formation. Nat Commun. 2016;7:10872. [46] Saleh KJ, Thongtrangan I, Schwarz EM. Osteolysis: medical and surgical approaches. Clin Orthop Relat Res. 2004:138-47. [47] Qu S, Bai Y, Liu X, Fu R, Duan K, Weng J. Study on in vitro release and cell response to alendronate sodium-loaded ultrahigh molecular weight polyethylene loaded with alendronate sodium wear particles to treat the particles-induced osteolysis. J Biomed Mater Res A. 2013;101:394-403. [48] Taki N, Tatro JM, Nalepka JL, Togawa D, Goldberg VM, Rimnac CM, et al. Polyethylene

and

titanium

particles

induce

osteolysis

lymphocyte-independent, mechanisms. J Orthop Res. 2005;23:376-83.

23

by

similar,

[49] von Knoch M, Jewison DE, Sibonga JD, Sprecher C, Morrey BF, Loer F, et al. The effectiveness of polyethylene versus titanium particles in inducing osteolysis in vivo. J Orthop Res. 2004;22:237-43.

Figure captions: Figure 1. TF3 suppressed particle-stimulated bone destruction in a mouse calvarial model. (A) Representative µCT images of mice with different treatment. Scar bar, 5mm. (B) BMD, (C) BV/TV, (D) porosity number and (E) area of porosity (%) were determined. (*P<0.05 and **P<0.01). Figure 2. Histological and histomorphometric analysis for the effects of TF3 on Ti-induced osteolysis. (A) Representative slices with H&E and TRAP staining. Scar bar, 100 µm. (B) The surface resorption area of the mouse calvariae (mm2), (C) periosteum thickness (mm), (D) TRAP-positive cells number and (E) OcS/BS were determined. (*P<0.05 and **P<0.01). Figure 3. TF3 inhibited osteoclastogensis induced by RANKL in a dose- and time-dependent manner in vitro without cytotoxicity. (A) The structure of TF3 with a molecular formula of C43H32O20 and a molecular weight of 868.7g/mol. (B) Bone-marrow-macrophages (BMMs) were incubated in induction medium (30ng/ml M-CSF (R&D Systems) and 50ng/ml RANKL (R&D Systems)) together with various concentrations of TF3 for 48h. The cell counting kit-8 (CCK-8, Dojindo, Kumamoto, Japan) assay were then performed to detect cell viability. TF3 treatment at concentration ≤ 10 µm had no effect on cell viability. (C) After pre-treatment with various concentration of TF3 (0.01, 0.1, 1µM) for 4h, BMMs were then incubated in induction medium together with various concentration of TF3 (0.01, 0.1, 1µM) for 5 days. Subsequently, TRAP staining was used to determine osteoclast formation and 24

differentiation. (D) The number and area of TRAP-positive cells containing ≧3 nuclei were counted. (E) BMMs were incubated in induction medium and 1µM TF3 were added at day 0, 2, or 4 respectively. After incubated for 5 days, the cells were fixed and stained for TRAP. (F) The number and area of TRAP-positive cells containing ≧3 nuclei were counted. The group untreated with TF3 was considered to be control. (*P< 0.05 and **P< 0.01). Figure 4. TF3 impaired osteoclastic bone resorption and formation of F-actin rings. (A) BMMs were plated on bone resorption plate and cultured with induction medium containing 30ng/ml M-CSF and

50ng/ml RANKL together with various

concentrations of TF3 for 5 days and the group untreated with TF3 was considered as control. Resorption pits were visualized at the magnification of 5× under an inverted microscope. (B)The area of bone resorption was determined by Image J 6.0 software. (*P<0.05 and **P<0.01). (C) After treated with or without TF3, BMMs were incubated with induction medium until the mature osteoclast were formatted and then stained

for

F-actin.

The

osteoclastic

rings

were

visualized

using

an

immunofluorescence microscope. Figure 5. TF3 suppressed the expression of RANKL-induced osteoclast-related genes including TRAP, NFATc1, c-Fos, CTSK and Oscar. (A) BMMs were incubated in 24-well plate in triplicate with induction medium with various concentration of TF3 for 5 days or (B) with 1µM TF3 for 1, 3 or 5 days. The gene copies of TRAP, NFATc1, c-Fos, CTSK and Oscar were quantified by RT-PCR. (*P< 0.05 and **P< 0.01). Figure 6. TF3 inhibited osteoclastogensis via ERK signaling pathway without affecting JNK, p38, NF-κB and PI3k/Akt signaling pathways. After presentative treatment with or without 1µM TF3 for 4h, RAW264.7 cells were stimulated by 25

50ng/ml RANKL for indicated times (0, 5, 15, 30, 60min) both in control (untreated with TF3) and TF3-treated groups. Then, the cells were collected and lysed for western blot analysis (A&C). The relative grey level of corresponding to AKT1, IκBα, p65, ERK, JNK and p38 phosphorylation were quantified by using Image J 6.0 software (B&D). (*P<0.05, **P<0.01). Figure 7. TF3 inhibited the activation of the ERK signaling pathway in a dose-dependent manner and downstream expression of c-Fos and NFATc1 without affecting upstream kinase of ERK. (A) After pre-treated with various concentrations of TF3 for 4h, RAW264.7 cells were stimulated with 50ng/ml RANKL for 15min. Cell lysates were subjected to western blotting against ERK antibody. (B) After treated with or without 1µM TF3 for 4h, RAW264.7 cells were stimulated with 50ng/ml RANKL for indicated times (0, 5, 15, 30, 60min). Cell lysates were then analyzed using western blotting against MEK1/2 antibody. (C) After treated with 1µM TF3 for 4h, BMMs were cultured in DMEM supplemented with 30ng/ml M-CSF and 50ng/ml RANKL for 0,1 or 3 days. Cell lysates were subjected to western blotting against c-Fos and NFATc1 antibodies. (D, E and F) The Image J 6.0 software was used to analyze the relative grey level of corresponding to ERK phosphorylation, MEK1/2 phosphorylation, c-Fos and NFATc1. (*P<0.05, **P<0.01).

26

27

28

29

30

31

32

33

34