Sotrastaurin, a PKC inhibitor, attenuates RANKL‐induced bone resorption and attenuates osteochondral pathologies associated with the development of OA

Abstract Osteoarthritis (OA) is a common degenerative disease that affects the musculoskeletal structure of the whole joint, which is characterized by progressive destruction of both articular cartilage and subchondral bone. Treatment of the bone pathologies, particularly osteoclast‐mediated subchondral bone loss in the early stages of OA, could prevent subsequent cartilage degeneration and progression of OA. In the present study, the PKC inhibitor, Sotrastaurin, was found to inhibit RANKL‐induced osteoclast formation in vitro in a dose‐ and time‐dependent manner. In particular, SO exerted its anti‐osteoclastic effect predominantly at the early stages of RANKL stimulation, suggesting inhibitory effects on precursor cell fusion. Using mature osteoclasts cultured on bovine bone discs, we showed that SO also exerts anti‐resorptive effects on mature osteoclasts bone resorptive function. Mechanistically, SO attenuates the early activation of the p38, ERK and JNK signalling pathways, leeding to impaired induction of crucial osteoclast transcription factors c‐Jun, c‐Fos and NFATc1. We also showed that SO treatment significantly inhibited the phosphorylation of PKCδ and MARCKS, an upstream regulator of cathepsin K secretion. Finally, in animal studies, SO significantly alleviates the osteochondral pathologies of subchondral bone destruction as well as articular cartilage degeneration following DMM‐induced OA, markedly improving OARSI scores. The reduced subchondral bone loss was associated with marked reductions in TRAP(+) osteoclasts in the subchondral bone tissue. Collectively, our data provide evidence for the protective effects of SO against OA by preventing aberrant subchondral bone and articular cartilage changes. Thus, SO demonstrates potential for further development as an alternative therapeutic option against OA.


| INTRODUC TI ON
Osteoarthritis (OA) is the most prevalent form of arthritis with a worldwide incident rate of over 70% and can be considered a leading cause of global disability. 1,2 OA is a debilitating joint disease that causes pain, joint stiffness and limited joint motion. 3,4 For many years, OA was considered primarily a cartilage disorder, but with advancement in technology and numerous studies, it is now recognized as a 'whole joint disease'. Due to the intimacy and interdependence of joint tissues, pathological changes in one joint tissue will ultimately compromise the structural integrity and function of other joint tissues. Pathological changes that occur at the osteochondral junction in OA include articular cartilage degeneration, subchondral bone loss and osteophyte formation. [5][6][7][8] Despite our knowledge of the osteochondral pathologies that occur during OA, the molecular underpinning that leads to these changes in OA is not fully understood; indeed, an enigma still remains as to whether articular degeneration precedes subchondral bone loss or vice versa. Mounting evidence suggest that increased osteoclast activity and high bone turnover rate leading to subchondral bone loss are an early instigator for subsequent cartilage degeneration in the development of OA. 6,9 Undeniably, the integrity of the articular cartilage depends on the underlying subchondral bone to function as a shock absorbers against high peak stresses. 10 Animal studies have shown in the early stages of OA, marked reduction in subchondral bone thickness in the as a result of elevated osteoclast activity leading to loss of structural integrity of the osteochondral junction and promoting cartilage degeneration. 11,12 Similarly, decreased connectivity of subchondral bone trabeculae has been observed in patients with early-stage OA which further reinforce the notion that subchondral bone loss is an early event that precedes cartilage damage and degeneration. These studies have provided compelling evidence for the pharmacological targeting of osteoclast bone resorptive activity in the early stages of OA for the treatment of OA. 13 Osteoclasts are unique and highly specialized multinucleated giant cells that possess the remarkable capacity to degrade mineralized bone and cartilage matrix. 14,15 Osteoclasts achieve this via the secretion of acid and bone lytic proteases such as TRAP and cathepsin K into the resorption pit which dissolves and degrades the underlying bone matrix, respectively. 16 For osteoclast to carry out bone resorption, extensive cytoskeletal polarization and formation of podosomal F-actin ring are essential. 17 This leads to the formation of the sealing zone and ruffled border membrane through which acid and proteolytic enzymes are released into the space between the osteoclasts and adjacent bone surface. 18 Protein kinase Cs (PKCs) are a family of serine/threonine kinases that play important roles in various cellular processes including cell proliferation, differentiation and survival. Of the numerous members, PKCδ has been found to be required for osteoclast bone resorption and survival. 19,20 In particular, PKCδ was shown to be necessary for ruffled border formation and cathepsin K secretion. 20 As such, targeting of PKCδ activity may offer potential therapeutic benefits in treating the bone pathologies associated with the development and progression of OA.
In this study, we examined the effects of Sotrastaurin (SO), a small molecular weight indolylmaleimide-based immunosuppressant that selectively and potently inhibits PKC, in osteoclast formation and bone resorption in vitro and potential in vivo benefits in murine model of surgical destabilization of the medial meniscus (DMM)induced experimental OA. Here, we found that SO inhibited osteoclast formation and bone resorption in vitro via the suppression of RANKL-induced activation of MAPK (p38, ERK and JNK) signalling cascades. Furthermore, these anti-osteoclastic and anti-resorptive effects of SO conveyed protective effects against DMM-induced subchondral bone loss and cartilage degeneration significantly improving OA outcomes.

| Ethics statement
The current experiment was authorized by approved by Zhejiang University Institutional Animal Care and Use Committee (No.12951).
All procedures related to animal use are carried out in accordance with the guidelines of the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals, and the maximum humane care for animals according to the guidelines approbated by Zhejiang University Institutional Animal Care and Use Committee. described. 21 Briefly, BMMs were isolated by flushing the marrow of excised femurs and tibias of 6 to 8-week-old C57BL/6 mice.
Upon confluence, adherent M-CSF-dependent BMMs were considered osteoclast precursor cells and used in subsequent downstream experiments.

| In vitro osteoclast differentiation assay
Bone marrow-derived macrophages were seeded into 96-well plate at density of 8 × 10 3 cells/well in complete α-MEM supplemented with 30 ng/mL M-CSF for 24 hours. To examine the dose-dependent effect of SO on osteoclast formation, cells were then stimulated with 100 ng/mL RANKL without or with increasing concentrations (1.5625, 3.125 or 6.25 μmol/L) of SO for 6 days. To assess the timedependent effect of SO on osteoclast differentiation, cells were stimulated with RANKL without or with 6.25 μmol/L of SO from day 0 to day 2 (D0-D2; early stage), day 2 to day 4 (D2 -D4; middle stage) or day 4 to day 6 (D4 -D6; late stage) of experimental period. Cells stimulated with RANKL only throughout the 6 day period in the absence of SO were used as controls. Media containing M-CSF, RANKL and SO were replaced every other day until typical well-spread multinucleated osteoclasts were observed in RANKLonly treated controls (around 6 days). At this point, cells were briefly and gently washed with phosphate-buffered saline (PBS), fixed in 4% paraformaldehyde (PFA) for 30 minutes and then stained for tartrate resistance acid phosphatase (TRAP) activity. The number of TRAPpositive multinucleated osteoclasts with three or more nuclei and the average cell-spread area were quantified using ImageJ software (NIH).

| Podosomal actin cytoskeleton immunofluorescence
BMM-derived osteoclasts were cultured as described above in the absence or presence of increasing concentrations (1.5625, 3.125 and 6.25 μmol/L) of SO. When mature well-spread osteoclasts were observed in RANKL-only treated controls (around 6 days), cells were fixed with 4% PFA for 30 minutes, washed briefly with PBS and then permeabilized with 0.1% Triton X-100 for 5 minutes at room temperature. Cells were then incubated with Actin-Stain 488 Phalloidin for 30 minutes in the dark at room temperature. The nuclei were counterstained with 4ʹ,6-diamidino-2-phenylindole (DAPI) for 5 mintues in the dark at room temperature. Fluorescence images were captured with the Eclipse TS100 Inverted Fluorescence Microscope (Nikon Instruments; Tokyo Japan). The number and average size of podosomal actin belt were quantified using ImageJ software.

| In vitro osteoclast-mediated bone resorption assay
M-CSF-dependent BMMs were stimulated with RANKL only for 3-4 days or until they form into small pre-osteoclasts and then transferred in the same amount onto the size 100 μm of sterilized bovine bone discs (Rongzhi Haida Biotech Co., Ltd), in triplicate. After adherence, cells were treated with 100 ng/mL RANKL without or with increasing concentrations (1.5625, 3.125 or 6.25 μmol/L) of SO.
Culture media was replaced every 2 days for further 6 days. At the end of the experimental period, cells were removed by mechanical agitation and processed for scanning electron microscopy imaging on a FEI Quanta 250 (Thermo Fisher Scientific). The average bone resorption area with respect to total bone disc area was quantified for each experimental condition using ImageJ software.

| Protein extraction and Western blot analyses
To assess the effects of SO on early RANKL-induced signal-

| Micro-CT scanning and analysis
The excised whole right knee joints from each experimental group were scanned using a μCT100 high-resolution cabinet cone beam

| Histological assessments
The scanned knee joints were decalcified in 10% EDTA (pH 7.4) for 21 days at 4°C and then embedded into paraffin blocks. . The amount and dimension of osteoclasts in each sotrastaurin concentration group were compared with the untreated controls. Data are expressed as means ± SD (*P < .05, **P < .01 and ***P < .001); SD, standard deviation. Scale bar, 100 μm. All experiments were performed at least three times

| Statistical analyses
The data presented in this study are expressed as mean ± standard deviation of at least three times independently conducted experiments.
Statistical difference was determined by Student's t test or one-way ANOVA with LSD test using SPSS 19.0 software (IBM). A P-value <.05 or unless otherwise indicated was considered statistically significant.

| Sotrastaurin inhibits RANKL-induced osteoclast formation in vitro
The potential cytotoxic effect of SO ( Figure 1A, chemical structure) was first assessed using the CCK-8 assay to identify sub-lethal con-  Figure 1C,D). Additionally, the osteoclasts that formed following SO treatment were significantly smaller ( Figure 1C,E) and less multinucleated ( Figure 1F) than untreated controls, suggesting an inhibitory effect on BMM precursor cell fusion.
To further determine at which stage of osteoclast formation SO exhibit its inhibitory effect, we stimulated BMM cells with RANKL and then treated with 6.25 μmol/L SO on specified days of osteoclast formation. As shown in Figure 1G,H, the strongest inhibitory effect was observed when cells were exposed to SO early on in the RANKL-induced osteoclast differentiation process between day 0 and day 2. Exposure of BMM cells within the first 2 days of RANKL stimulation markedly reduced their ability to differentiate and fuse into well-spread multinucleated osteoclasts ( Figure 1I,J). In contrast, when BMM cells were exposed to SO later in the differentiation process, that is between day 4 and day 6, only a moderate reduction in osteoclast formation was observed. Well-spread TRAP(+) multi-

| Sotrastaurin impairs bone resorption in vitro
Having now shown that SO inhibits RANKL-induced osteoclast formation via the impairment of precursor cell fusion, we sought to determine the effects of SO on mature osteoclast bone resorptive function. Given that SO inhibits the formation of the podosomal actin belt which is also required for bone resorption, we hypothesize F I G U R E 2 Sotrastaurin inhibits podosome actin belt formation and impairs osteoclasts mediated bone resorption in vitro. A, Effect of sotrastaurin on the formation of podosome actin belt. M-CSF-dependent BMMS were seeded onto 96 well plates and stimulated with RANKL without or with the specified concentration of sotrastaurin (0, 1.625, 3.125 and 6.25 μmol/L) for 6 d. Cells were fixed and stained for immunofluorescence. Typical fluorescence images of osteoclasts' podosome actin belts (green) and nuclei (blue, DAPI). B and C, The number and area of podosome actin belts were quantified. Scale bar, 100 μm. D, Effect of sotrastaurin on osteoclast bone resorption activity. BMM-derived pre-osteoclasts were cultured on bone slices and stimulated with RANKL (100 ng/mL) in the absence or presence of increasing concentrations of sotrastaurin for 6 d. The cells were removed by ultrasound and the resorption pits were evaluated under a scanning electron microscope. Scale bar, 100 μm. E, Relative to the untreated controls, area of bone resorption pits under each treatment concentration was quantified. F, BMMs were treated with 6.25 μmol/L sotrastaurin, cells were incubated for 24 and 48 h and then harvested. The expression levels of phosphorylated PCKδ and MARCKS in protein lysates were analysed by Western blot. Expression of β-Actin was analysed as an internal loading control. G, Densitometric analyses of the expression of the phosphorylation-PKCδ and phosphorylation-MARCKS were reported as p-PKCδ/ β-Actin and p-MARCKS/ β-Actin. Data are presented as the mean ± SD (*P < .05, **P < .01 and ***P < .001); SD: standard deviation. DAPI: 6-diamidino-2-phenylindole that SO will inhibit mature osteoclast bone resorption in vitro. To this end, BMM-derived osteoclasts cultured on bovine discs were stimulated with indicated concentrations of SO for 6 days after which cells were removed and resorption pits analysed. As shown in the scanning electron micrographs in Figure 2D,E, osteoclasts cultured without SO treatment displayed significant bone resorptive activity, resorbing close to 80% of the total bone disc area. On the other hand, osteoclasts treated with SO showed marked reductions in bone resorption capability ( Figure 2D), resorbing only 40%, 10% or 5% of total bone disc area when treated with 1.5625, 3.125 F I G U R E 3 Sotrastaurin exerts its anti-osteoclast effect by the inhibition of MAPKs signalling cascade. A, Sotrastaurin inhibited the activation phosphorylation of p38, JNK and ERK induced by RANKL. After pre-treatment with 6.25 μmol/L sotrastaurin for 1 h, and then stimulated with RANKL for a specified time, Western blot analysis was performed with specific antibodies to mitogen-activated protein kinases (p38, ERK and JNK) signalling cascade. β-actin was used to act as an internal loading control. B, Relative change in the phosphorylation status of p38, JNK and ERK were determined by optical density analysis of each phosphorylated band, and each one was expressed as the ratio of its total protein counterpart. C, Sotrastaurin attenuated the expression of c-Fos, c-Jun and NFATc1 induced by RANKL. The total protein extracts of BMMs cultured in the medium containing RANKL and 6.25 μmol/L sotrastaurin for 0, 1, 3 and 5 d were analysed by Western blotting and specific antibodies to c-Fos, c-Jun and NFATc1. β-actin was used to act as an internal loading control. D, The changes in protein expression level after sotrastaurin treatment were quantified by relative β-actin density analysis and expressed as the ratio compared with the untreated group stimulated by RANKL alone. E, Sotrastaurin inhibited osteoclast marker genes induced by RANKL in a dose-dependent manner. Real-time quantitative PCR assessed the expression of PKCδ, ACP5, NFATc1, DC-STAMP, c-Fos and CTSK in osteoclasts treated with specified concentrations of sotrastaurin. The expression of the target genes were standardized with reference to the housekeeping gene GADPH and then displayed a decreased change compared with the untreated control groups. Data are presented as the mean ± SD (*P < .05, **P < .01 and ***P < .001); SD, standard deviation. MARCKs, myristoylated alanine-rich C-kinase substrate or 6.25 μmol/L SO, respectively ( Figure 2E). Previous studies have shown that PKCδ regulates cathepsin K secretion during osteoclast bone resorption via the modulation of myristoylated alanine-rich C-kinase substrate (MARCKS). 20 Here, we further confirmed that the inhibition of osteoclast bone resorption activity by SO was due to inhibition of PKCδ and MARCKS phosphorylation ( Figure 2F,G).
Thus, these results further show that SO also exhibits anti-resorptive effects against mature osteoclast bone resorption in vitro.

| Sotrastaurin attenuated RANKL-induced activation of MAPKs signalling cascade
The MAPK signalling pathway consisting of members, p38, ERK and JNK, is intimately involved in OA-related cartilage destruction as well as osteoclast formation and bone resorption. [24][25][26] Furthermore, PKCδ has been shown to be involved in the upstream regulation of MAPK signalling in response to RANKL-RANK activation. 19 Thus, we further investigated the effects of SO on RANKL-induced signalling events. Binding of RANLK to receptor RANK rapidly induces the activation phosphorylation of all three members of MAPK signalling cascade. As shown in Figure 3A,B, p38, ERK and JNK phosphorylation was observed within 5 minutes of RANKL stimulation and lasting for around 15 minutes before returning to basal levels.
In contrast, pre-treatment of cells with SO significantly attenuated the activation phosphorylation of all three members of the MAPK signalling cascade (Figure 3A Figure 5A,B, female mice; Figure S2a,b, male mice). On the other hand, the severity of cartilage degeneration and destruction in the SO treatment group was much less than the DMM-OA group, with the articular cartilage surface in the SO treated group exhibiting noticeably more surface regularity. Furthermore, HC thinning and CC thickening were not as pronounced as seen in the DMM-OA group ( Figure 5A,B, female mice; Figure S2a,b, male mice). The histological findings were supported by OARSI scores which were significantly increased in DMM-OA group and markedly reduced following SO treatment ( Figure 5C, female mice; Figure S2c  Pamidronate also protected Runx2 transgenic mice (mice with high bone remodelling) from bone loss associated with partial medial meniscectomy-induced OA. 38 Moreover, risedronate treatment in patients was found to reduce the level of C-terminal crosslinking telopeptide of type II collagen (CTX-II), a marker of cartilage degradation associated with progressive OA, but sustained clinical improvements in disease progression were not observed. 39 Thus, these studies provide evidence that targeting osteoclast formation and activity to prevent subchondral bone loss to maintain articular cartilage integrity may offer beneficial symptomatic and structural benefits against OA development and progression. 23 Osteoclasts are multinucleated giant cells derived from the fu-  Thus, it is likely that SO will exert anti-inflammatory properties that can contribute to the overall protective effects against OA.

| Sotrastaurin treatment averts subchondral bone deterioration and articular cartilage degeneration in mice with DMM-induced OA
In conclusion, our study demonstrated the beneficial role of PKC inhibitor, Sotrastaurin, in the treatment of OA. SO can inhibit RANKL-induced osteoclast formation and bone resorption in vitro and protects against DMM-induced OA osteochondral pathologies in vivo. Thus, SO not only prevents subchondral bone loss but also maintain the structural integrity of the overlying articular cartilage which substantially hinders the progression and severity of OA.
Together, our data suggest that Sotrastaurin may be further developed as a potential pharmacological agent for the treatment of OA.

ACK N OWLED G EM ENTS
Not applicable.

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interests to declare.

AUTH O R CO NTR I B UTI O N S
LS and PC designed the experiments and drafted the manuscript.
PC, WL and LX conducted the in vivo animal surgery and experiments. PC, WL and QH carried out the in vitro cellular and biochemical experiments. PC, WL and ZB analysed and interpreted the results. All authors reviewed the manuscript prior to submission.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this experiment are available from the corresponding author upon reasonable request.