The Potential Role of Wedololactone in Treating Particle-induced Osteolysis _A Mouse Calvarial Model

Background: Osteolysis is one of the most prevalent clinical complications of total joint replacement (TJR). Wedelolactone (WDL) is a coumestan compound derived from the Wedelia chinensis plant and has been demonstrated to exhibit anti-inammatory properties. This study aimed to investigate the use of WDL as a potential treatment for reducing the risk of particle-induced osteolysis using a well-established particle-induced mice calvarial disease model. Methods: Thirty-two C57BL/6J mice were randomized into four groups: sham, polystyrene particles (PS), PS particles with WDL treatment for 4 weeks (WDL 4w) and PS particles with WDL treatment for 8 weeks (WDL 8w). Micro-CT was used to quantitatively analyze the bone mass. Osteoclast numbers were also measured from histological analysis. Results: The results showed that bone mineral density was signicantly higher in the WDL 8w group than in the PS group (p < 0.05), and both the WDL 4 and WDL 8w groups had lower osteoclast numbers (p < 0.05). No signicant difference in osteoclast number was found between the WDL 4w and WDL 8w groups. Conclusions: These results support the use of WDL as a herbal medicine for reducing the severity of particle-induced osteolysis after TJR.


Background
In clinical orthopaedic practice, arti cial joint replacement (AJR) is considered an effective method for treating severe joint degeneration [1]. However, periprosthetic osteolysis resulting from the deposition of wear particles from the articulating joint is one of the major long-term complications following AJR [2].
The wear particles stimulate in ammatory responses and osteoclastic resorption processes at the bone implant interface, which consequently often leads to implant loosening [3,4]. Although bearing materials have been introduced to reduce the generation of wear particles, osteolysis is still prevalent and is considered a major long-term complication of AJR.
Besides changing the base material of the implant, pharmaceuticals have also been investigated as a method for reducing osteolysis [5]. Bisphosphonates are well-known drugs for treating osteoporosis, but have also been shown to be effective at suppressing osteolysis [6,7]. Similarly, statins, which are lipidlowering agents, have been reported to reduce particle-induced osteolysis in a murine calvarial model [7].
However, such drugs can also present considerable serious side effects, such as atypical femoral bone fracture and osteonecrosis of the jaw [8][9][10].
Animal models are often used for investigating mechanisms that can lead to particle-induced osteolysis and for evaluating suitable treatment methods [7,[11][12][13][14]. A previous study by our institute investigated whether strontium ranelate [12], a drug for osteoporosis, could be effectively used to combat osteolysis.
After gavage-feeding mice for up to 4 weeks, the results showed a signi cant increase in bone mineral density (BMD), bone volume/tissue volume (BV/TV) and trabecular thickness (Tb.Th), and a signi cant reduction in osteoclast numbers [12]. However, long-term use of SR presents could raise a risk of cardiovascular disease. [15][16][17].
As an alternative to pharmaceuticals, Chinese herbal medicines are considered a more natural solution with fewer side effects [18]. In addition, the International Organization for Standardization (ISO/TC 249) has formulated a series of standards of herbal medicines, and some studies have started to pay attention to the bene t in disease treatments [19][20][21][22]. Wedelolactone (WDL) is a coumestan compound extracted from the Wedelia Chinensis plant [23]. WDL is considered a traditional Chinese herbal medicine with strong anti-in ammatory properties [24][25][26][27][28]. Recent studies have shown that WDL can help promote hair growth [29] and is hepatoprotective [30,31], neuroprotective [32] and anti-carcinogenic [33,34].
Few studies have reported on the potential of WDL for treating diseases of the skeletal system. Based on the ability of WDL to reduce in ammation and inhibit osteoclastogenesis, it is hypothesized that WDL could also play a role in reducing the risk of particle-induced osteolysis. This study used a wellestablished murine calvarial osteolysis model to investigate whether WDL administered orally could reduce the severity of osteolysis.

Methods
Establish the calvarial particle-induced osteolysis animal model The protocol for this experiment was approved by the Institutional Animal Care and Use Committee at the institute where the study was performed. Thirty-two 6 week old C57BL/6J female mice were supplied by BioLASCO (Taipei, Taiwan), an AAALAC certi ed biotechnology company. The animals were kept in a room at 24℃, 50% humidity, and with a 12h light/dark cycle (light from AM 7:00 to PM 7:00). Animals were separated at random into four groups: (1) sham group (n = 8) (underwent surgery only) (2) vehicle group (n = 8) (implanted with PS particles), (3) WDL 4w (n = 8, PS particles and treated with WDL for 4 weeks. 7 animals remained at the end of the experiment) and (4) WDL 8w (n = 8, treatment for 8 weeks). Polystyrene (PS) particles have been proven to induce osteolysis (35), and so the vehicle group and WDLtreated groups were injected with 1mg PS particles/100 µl HA [11,26]. The polystyrene particles (Polystyrene Latex Spheres, 610 − 38) were purchased from TED PELLA, Inc. (CA, USA). Three hundred particles were randomly selected and measured through SEM to measure the particle size and aspect ratio (Fig. 1A). The particles were found to be 1.03 ± 0.04 µm and 0.99 ± 0.03, respectively (Fig. 1B, C).
The particles were also tested to have an endotoxin level below 0.25 EU/mL using a Limulus Ambocyte Lysate assay kit (ToxinSensor™ gel clot endotoxin assay kit, GenScript, NJ, USA) and then suspended in hyaluronic acid.
To implant the PS particles, the mice were anesthetized with 100 mg/kg of Zoletil 50 and 10 mg/kg Rompun by intraperitoneal injection. A 0.5 × 0.5 cm area of the middle calvaria was exposed by sagittal incision. After removing the periosteum intact, the PS particle suspensions were spread over the area and the incision was closed with sutures. After 2 weeks post-surgery, the WDL 4w and 8w groups were gavage-fed with Wedelolactone (Y0001599, European Pharmacopoeia Reference Standard, Sigma-Aldrich, USA) at a dose of 4 mg/kg/day for 5 days/week [21]. WDL was dissolved by DMSO as stock solution and diluted with phosphate-buffered saline (PBS) into 10 volumes to attain a dosage of 4 mg/kg. The vehicle group was gavage-fed the vehicle solution (10% DMSO in PBS). The sham group, vehicle group and WDL 4w group were then sacri ced after 4 weeks of feeding vehicle or WDL, and the WDL 8w group was sacri ced after 8 weeks of feeding WDL.

Micro-CT imaging analysis
The calvarias were xed in 10% buffered formalin for 24 h, and then transferred to 70% ethanol for 24h. The specimens were scanned with the micro-CT system Skyscan 1076 (Bruker micro-CT, Kontich, Belgium) at a resolution of 2048×2048. Three-dimensional images were reconstructed in Skyscan system with a voxel size of 9 µm [11,26]. A spherical volume of interest (VOI) of 5 mm in diameter was then de ned with the bregma as the center. Within this VOI, the bone mineral density (BMD, mg/cc), the ratio of bone volume to tissue volume (BV/TV, %) and trabecular thickness (Tb.Th) were recorded for each group.

Histological analysis
The calvarias were decalci cated in 10% ethylenediaminetetraacetic acid (EDTA) for 2 weeks, and para n embedding. Five µm thick sections of the calvaria were taken in the sagittal plane centered over the particle-treated area. The sections were then stained with hematoxylin and eosin (H&E stain) to observe the morphology of in ammatory response of the connective tissue. A tartrate-resistant acid phosphatase (TRAP) stain was performed using a commercial TRAP kit (#386A, Sigma-Aldrich). The number of osteoclasts was determined by counting the number of TRAP-positive multinucleated cells by two coauthors to eliminate intra-and inter-observer error.

Statistical analysis
The data were analyzed by one-way analysis of variance (ANOVA) to show the difference between groups. Multiple comparisons were adjusted with a Bonferroni post hoc test. Results were reported as mean ± standard deviation (SD). p-values of less than 0.05 were considered signi cantly different.

Micro-CT imaging analysis
A visual analysis of the three dimensional (3D) reconstructed micro-CT images showed clear differences between the sham group, vehicle group, and WDL groups ( Fig. 2A). The images show typical osteolysis with pores in the sham group, but both the size and number of pores decreased in WDL-treated groups.
The presence of PS particles signi cantly decreased the BMD in the vehicle group by 7.8% when compared to the sham group (0.744 ± 0.03 for vehicle group and 0.801 ± 0.03 for sham group, p < 0.01). Both WDL groups showed an increase in BMD, with the 8w group showing a signi cant increase of 5.1% in comparison to the vehicle group (0.784 ± 0.01 for 8w group and 0.744 ± 0.03 for vehicle group, p < 0.05). The BV/TV in the vehicle group decreased by 4.1% in comparison to the sham group (21.23 ± 1.43 in vehicle group versus 22.1 ± 1.54 in sham group) but increased in the WDL 4w group (22.96 ± 2.07) and WDL 8w group (22.44 ± 2.78). There was no signi cant difference in BV/TV between the WDL 4w and WDL 8w. There was also no signi cant difference in Tb.Th between all four groups (Fig. 2B).

Histomorphometric analysis
A H&E stain was used to evaluate the in ammatory response through histological analysis. Pseudomembrane proliferation occurred in the vehicle and WDL groups. The morphology of the cells in the periosteum was observed to change to a circle-shaped contour, while the cells resembled a at contour in the sham group. Multinucleated giant cells were found in the surrounding periosteum (Fig. 3). TRAP staining was used to highlight PS particles in the periosteal cells and multinucleated giant cells ( Fig. 4A & B).
Osteoclasts around the bone perimeter A TRAP stain was used to highlight osteoclasts around the calvaria in order to calculate the osteoclast numbers in each group. The results showed the osteoclast numbers in the vehicle group increased signi cantly in comparison to the sham group (43.74 ± 10.09 in vehicle group versus 18.25 ± 14.67 in sham group, p < 0.05), demonstrating that the polymer particles likely induced osteolysis. Furthermore, there was a signi cant reduction in osteoclast numbers in the WDL 4w group in comparison to the vehicle group (21.09 ± 9.79 in WDL 4w group versus 43.74 ± 10.09 in vehicle group, p < 0.05). No signi cant difference in osteoclast numbers was found between the WDL 4w and WDL 8w groups (30.56 ± 3.98) (Fig. 5).

Discussion
Osteolysis is one of the major long-term complications affecting patients following AJR. Chinese herbal medicine is considered to be milder than pharmaceutical treatments and does not produce strong adverse effects. This study aimed to investigate the potential use of the Chinese herbal medicine wedelolactone for reducing the incidence of particle-induced osteolysis.
Bisphosphonates (BPs) are commonly used to treat conditions of metabolic bone loss such as osteoporosis [36], but have also been shown to inhibit particle-induced osteolysis (37). However, the longterm use of BPs and other pharmaceuticals can often result in serious adverse effects, such as increasing the risk of osteonecrosis of the jaw, atypical femur fractures, atrial brillation, and esophageal cancer [38]. Statins, a drug usually used to lower blood cholesterol levels and reduce symptoms risk for illnesses related to atherosclerosis, targets the mevalonate pathway of osteoclasts, which affects the same inhibition mechanism as bisphosphonates. Statins have been shown to markedly reduce the severity of particle-induced osteolysis in a murine calvarial model [7]. However, as with BPs, the use of statins can present a number of side effects when used long-term, such as rhabdomyolysis, cognitive loss, neuropathy, hepatic dysfunction, and sexual dysfunction [9].
There is increasing interest in alternative methods like Traditional Chinese Medicine for treating disease as clinicians look to reduce long-term complications associated with conventional medicine. It has reported that postmenopausal Chinese women with greater fruit intake have a signi cantly higher BMD than comparable women with a lower fruit intake [39]. Flavonoids, found in a wide diversity of plant foods from fruits, have the most potential of dietary components for promotion of bone health than general fruits and vegetables consumption [40,41]. Phytoestrogens, which are natural compounds that act to maintain healthy bones, have been shown to protect against postmenopausal bone loss [20]. This protective mechanism has been demonstrated with avones [42][43][44], avanones [45], avonols [46], coumestans [47], and triterpenoids [45,48]. Some phytoestrogens also have the ability to reduce osteolysis by blocking some modules in the RANKL signaling pathway, and subsequently reducing the release of cytokines [42][43][44].
To our best knowledge, no studies to date have investigated whether WDL could reduce particle-induced osteolysis using an in-vivo mice calvarial model. The concentration of WDL in the current study was adopted from Tsai et al. [21], who showed that a low oral dose of WDL (4mg/kg) for 4 weeks signi cantly suppressed the growth of prostate cancer cells. The results of our study showed that the BMD was signi cantly greater in the WDL 8w group (0.784 ± 0.014 mg/cc) than in the vehicle group (0.744 ± 0.032 mg/cc) (Fig. 2B). There was no signi cant difference between the WDL 4w and 8w groups in terms of the mean BV/TV Tb.Th, but both values were signi cantly less in the vehicle group. On the other hand, the osteoclast numbers were signi cantly lower in the WDL 4w group (21.09 ± 9.79) and WDL 8w group (30.56 ± 3.98) than the vehicle group (44.09 ± 9.83) (Fig. 5).
Previous studies treated murine calvarial osteolysis with bioactive compounds for 10 to 14 days after implantation of foreign particles [13, 42-44, 48, 49]. For instance, icariin, a bioactive avonoid, has been proven to inhibit postmenopausal osteoporosis. Shao et al. gavage-fed mice with icariin at doses of 0.1 mg/g and 0.3 mg/g for 14 days to examine the effects on osteolysis in a particle-induced murine calvarial model. The results showed an increase in BMD and BV/TV over the control model, and the number of TRAP positive cells decreased [13]. Similarly, ursolic acid is an abundant triterpenoid present in over one hundred species of plants. It has been reported that ursolic acid isolated from loquat leaves could reduce bone loss in OVX mice [45]. Jiang et al. treated mice with 10 mg/kg and 40 mg/kg doses of ursolic acid of intraperitoneal injection for 14 days that ursolic acid protects against wear particleinduced osteolysis by suppressing osteoclast formation and function [48]. Compared with these researches, this current study fed mice an oral dose of WDL (4 mg/kg) for 4 and 8 weeks. The treatment period in this study was longer than the other studies referenced above treating for only 2 weeks which was considered a short-term treatment. No adverse effects were observed in this study after treating the mice for 8 weeks with WDL.
Bone remodeling is a dynamic equilibrium. Although the trigger mechanisms for osteolysis are not yet fully understood, it is known that one of the mechanisms is the receptor activation of NF-κB ligand (RANKL) and osteoprotegerin (OPG) secreted from osteoblasts and osteogenic stromal cells, both of which act to maintain the bone balance when acting normally [50]. RANKL is required for the differentiation of osteoclast precursors into mature osteoclasts [51]. As the ratio of RANKL/OPG increases, the osteoclast precursors are easier in uenced by RANKL signaling through the downstream activation of NF-κB/c-fos/NFATc1, subsequently causing the precursors to differentiate into mature osteoclasts. On the other hand, macrophages also plays a key role in wear particle-induced osteolysis. Cytokines (TNF-α and IL-1β, etc.) and other mediators of pro-in ammation from activated macrophages can regulate or stimulate other tissue-resident macrophages to promote osteoclastogenesis. These cytokines also regulate JNK and the p38/Erk signaling pathway to induce NFATc1, one of the downstream factors in the RANKL signaling pathway, which can lead to osteolysis.
Studies have shown that some compounds from Chinese herbal medicines can treat particle-induced osteolysis by inhibiting the modules in the NF-κB signaling pathway, the main osteolysis-related mechanism, to effect the balance of osteoclasts and osteoblasts [13,44,48,49]. WDL is known for its ability to block the phoporylation of IκBα, which acts to regulate the transcription of NF-κB mediated genes, inhibiting LPS-induced pro-in ammation [49]. Annie et al. demonstrated how an extract from Wedelia chinensis attenuated Ovariectomy (OVX)-induced bone loss in mice [52]. WDL extracted from Ecliptae herba has been shown to inhibit osteoclastogenesis of RAW 264.7 cells treated with RANKL [53], and prevent OVX-induced bone loss by inhibiting osteoclast activity and enhacing osteoblast activity [54]. Furthermore, it has been con rmed WDL regulated RANKL-related NF-κB/c-fos/NFATc1 pathway can suppress osteoclastogenesis [27,55], and also regulated Wnt/β-catenin signaling pathway to induce osteoblastogenesis [55,56]. The authors concluded that oral WDL could improve bone formation and inhibit resorption by affecting the balance of osteoclasts and osteoblasts. However, the mechanism leading to the inhibition of osteolysis by WDL still needs to be determined.
Some limitations of this study should be mentioned. First, murine calvarial models allow for a low-cost study with relatively quick results, but the models use a at bone instead of a long bone and the particles are injected on the cortical bone surface rather than in cancellous bone. Second, as detailed above, the WDL dose used in this study were adopted from other related publications. But the most effective dose for treating osteolysis in vivo has yet to be determined and requires further study. Third, osteoclast numbers were counted through qualitative analysis, not quantitative analysis. When injected onto the calvaria, the particles randomly precipitated and then a section was chosen for histological staining. This sampling approach may not represent realistic results of osteoclast numbers. Accepting the above limitations, this animal study identi ed the potential role of wedelolactone for treating particle-induced osteolysis.

Conclusion
This study indicated that wedelolactone (WDL), a Chinese herbal medicine, could help to maintain bone quality. Oral WDL was shown to suppress osteoclast numbers and maintain the level of BMD over time in this particle-induced osteolysis murine clavarial model. Moving forward, WDL could be potentially developed as a functional food for lowering the risk of particle-induced osteolysis after total joint replacement.

Availability of data and materials
The data used and/or analyzed during the current study are available from the corresponding author on reasonable request SEM image (×12,000) of polystyrene (PS) particles (A). The particle size distribution by light-scattering analysis (B). The particle size distribution (C) and the aspect ratio (D) of the particles from SEM images.

Figure 2
Reconstructed image of the VOI with the bregma at the center. The VOI is de ned with a diameter of 5 mm (A). Micro-CT image of bone formation in a particle-induced osteolysis model measured at 4 and 8 weeks after feeding WDL (B) (*p < 0.05; ** p < 0.01, as determined using ANOVA testing).