The Fungal Metabolite (+)-Terrein Abrogates Ovariectomy-Induced Bone Loss and Receptor Activator of Nuclear Factor-κB Ligand–Induced Osteoclastogenesis by Suppressing Protein Kinase-C α/βII Phosphorylation

Osteoporosis is a common disease characterized by a systemic impairment of bone mass and microarchitecture that results in fragility fractures. Severe bone loss due to osteoporosis triggers pathological fractures and consequently decreases the daily life activity and quality of life. Therefore, prevention of osteoporosis has become an important issue to be addressed. We have reported that the fungal secondary metabolite (+)-terrein (TER), a natural compound derived from Aspergillus terreus, has shown receptor activator of nuclear factor-κB ligand (RANKL)–induced osteoclast differentiation by suppressing nuclear factor of activated T-cell 1 (NFATc1) expression, a master regulator of osteoclastogenesis. TER has been shown to possess extensive biological and pharmacological benefits; however, its effects on bone metabolism remain unclear. In this study, we investigated the effects of TER on the femoral bone metabolism using a mouse-ovariectomized osteoporosis model (OVX mice) and then on RANKL signal transduction using mouse bone marrow macrophages (mBMMs). In vivo administration of TER significantly improved bone density, bone mass, and trabecular number in OVX mice (p < 0.01). In addition, TER suppressed TRAP and cathepsin-K expression in the tissue sections of OVX mice (p < 0.01). In an in vitro study, TER suppressed RANKL-induced phosphorylation of PKCα/βII, which is involved in the expression of NFATc1 (p < 0.05). The PKC inhibitor, GF109203X, also inhibited RANKL-induced osteoclastogenesis in mBMMs as well as TER. In addition, TER suppressed the expression of osteoclastogenesis-related genes, such as Ocstamp, Dcstamp, Calcr, Atp6v0d2, Oscar, and Itgb3 (p < 0.01). These results provide promising evidence for the potential therapeutic application of TER as a novel treatment compound against osteoporosis.


INTRODUCTION
At present, more than 15 million people have osteoporosis, which is a serious problem in Japan, where the population is extremely aged (Hagino et al., 2021). In the United States, there were more than 2 million osteoporosis-related fractures in 2005, with associated costs of approximately $ 17 billion. By 2023, the annual number of fractures will exceed 3 million, and the U.S. healthcare system is projected to cost $ 25 billion (Burge et al., 2007). Osteoporosis is caused by a deficiency of female hormone estrogen, which breaks bone homeostasis between osteoclasts and osteoblasts. Severe bone loss due to osteoporosis triggers pathological fractures and, consequently, decreases daily life activity and the quality of life. Therefore, the prevention of osteoporosis is one of the most urgent issues in Japan (Tilman et al., 2011;Katrin, 2019).
Currently, bisphosphonates (BPs) are mainly used in the treatment of osteoporosis (Shibahara et al., 2018). However, medicine-induced osteonecrosis of the jaw (MRONJ), which occurs after invasive dental procedures, such as tooth extraction, scaling, and root planning, has become a major problem in the treatment of osteoporosis, including BP treatment (Yoneda et al., 2017). The incidence of MRONJ in Japan is 0.85 per 100,000 people per year (Yoneda et al., 2017). According to a nationwide survey by the Japanese Society of Oral and Maxillofacial Surgery, it is estimated to be approximately 0.01-0.02% (Yoneda et al., 2010). In Japan, the number of remaining teeth in the elderly continues to increase as a result of the 8,020 movement (Nasu and Nakamura, 2016). Therefore, since MRONJ significantly decreases the quality of life of patients, it is desirable to develop new osteoporosis drugs without serious side effects.
Our research group has identified (+)-terrein (TER, Figure 1) as a naturally occurring compound with a different mechanism of action from that of molecularly targeted drugs, which is expected to have an anti-inflammatory effect by inhibiting interleukin (IL)-6 (Mandai et al., 2014). TER is a small molecule with a molecular weight of 154.16 kDa isolated from Aspergillus terreus as a secondary metabolite (Raistrick and Smith, 1935). TER has been reported to have various biological effects, such as inhibiting biofilm formation (Kim et al., 2018), angiogenin secretion in prostate cancer cells (Arakawa et al., 2008), and pulp inflammation (Lee et al., 2008). In addition, our group has established a route for the organic synthesis of TER (Mandai et al., 2014). Furthermore, this synthesized TER inhibits IL-6-induced vascular endothelial growth factor (VEGF) secretion in human gingival fibroblasts (Mandai et al., 2014). In addition, synthetic TER suppressed IL-6-induced macrophage colony-stimulating factor (M-CSF) secretion by inhibiting Janus-activated kinase (JAK)-1 phosphorylation (Yamamoto et al., 2018). We have also reported that TER has an inhibitory effect on RANKL or tumor necrosis factor (TNF)-α-induced osteoclastogenesis by suppressing the expression of nuclear factor of activated T-cell c1 (NFATc1), a master regulator of osteoclast differentiation in mouse bone marrow macrophages (mBMMs) (Nakagawa et al., 2020). In addition, we also showed that TER did not suppress RANKL-induced phosphorylation of NF-κB and MAPKs (ERK1/ 2 and p38), which are related to NFATc1 expression, meaning which it is still unknown about the effects of TER on RANKL signaling (Nakagawa et al., 2020). Based on these in vitro findings, we hypothesized that TER could suppress bone resorption in the femur if its anti-inflammatory and anti-osteoclastogenic effects were effectively exerted in the bones. However, the in vivo effect of TER on bone destruction remains unknown. In addition, to the best of our knowledge, there have been no studies on whether TER also acts on osteoblasts and whether it has an effect on bone homeostasis.
In this study, we evaluated the effect of TER on bone homeostasis in vivo using a mouse ovariectomy model of osteoporosis (OVX mice). We also assessed the effects of TER on the signaling pathways of RANKL-induced osteoclast differentiation using mBMMs in vitro, especially focused on protein kinase-C cascade which is also related to NFATc1 expression (Shin et al., 2014). PKC family members phosphorylate a wide variety of protein targets and are known to be involved in diverse cellular signaling pathways. In RANKLinduced osteoclast formation, PKC β functions, but not PKCδ, occur through glycogen synthase kinase-3β (GSK-3β)-NFATc1 axis (Shin et al., 2014). We also assessed the effects of TER on osteoblast differentiation using MC3T3-E1 cells in vitro. Our results suggest that host modulation using TER could be a novel therapeutic modality for the prevention of bone resorption, including osteoporosis.

The Ovariectomized Mice Model
Female C57BL/6J mice were obtained from Japan CLEA (Japan) and allowed to acclimatize for 1 week. At 8 weeks of age, mice were randomly divided into four groups (n 6-7 per group): 1) Sham, 2) OVX, 3) OVX + TER (10-30 mg/kg, two times a week), and 4) OVX + ZOL (0.1 mg/kg, two times a week). A fake operation was performed in the Sham group as a control, and the other three groups underwent bilateral OVX under anesthesia with sodium pentobarbital (60 mg/kg) via intraperitoneal injection (Thompson et al., 1995;Xiao et al., 2018;Xu et al., 2018). The Sham and OVX groups received equal doses of phosphate-buffered saline (PBS) as controls, the OVX + TER group was given an intraperitoneal injection of TER at the indicated concentrations, and the OVX + ZOL group received an intraperitoneal injection of ZOL at the indicated concentrations after surgery. After 8 weeks, all mice were euthanized via cardiac stick exsanguination. Serum samples were prepared and stored at −80°C for the measurement of tumor necrosis factor (TNF)-α, IL-1β, IL-6, and M-CSF. The femurs of the mice were collected. This animal experiment was performed in accordance with the Guidelines for Proper Conduct of Animal Experiments of the Science Councils of Japan and approved by the Animal Research Control Committee of Okayama University (approval no: OKU-2018756), and the mice were kept under SPF conditions ( Figure 2).

Image Analysis Using Micro-Computed Tomography
To observe the femur bone resorption, 4-μm-thick slices from the mice from each group were photographed using an animal micro-CT imaging device (Skyscan: Bruker, Kontich, Belgium), and these images were reconstructed using the accompanying software (Reconstruction software, NRecon: Bruker) and analyzed using the image processing software (Analyze software, CT-Analyzer: Bruker). We measured the indicators of bone mineral density (BMD, mg/cm 3 ), bone volume/trabecular volume (BV/TV, %), trabecular number (Tb.N, 1/μm), and trabecular separation (Tb.Sp, μm).

Histological Analysis
The femur samples were fixed in a 4% paraformaldehyde solution (pH 7.4; Wako Pure Chemicals, Osaka, Japan) for 1 day. The blocks were immersed in disodium ethylenediaminetetraacetic acid (10% EDTA 2Na solution, pH 7.0, Muto Chemical Co., Ltd., Tokyo, Japan) for 10 days for demineralization, and were dehydrated using an ethanol series. Paraffin-embedded blocks were sliced into 4-μm-thick sections. Subsequently, histopathological analysis was performed using the following staining methods.

Hematoxylin-Eosin Staining
The prepared paraffin sections were subjected to HE staining, according to the normal method, and then encapsulated using a cover glass and Mount-Quick (Daido Sangyo Co., Ltd., Tokyo, Japan). Histological images were then observed under an optical microscope (BX-50; Olympus, Osaka, Japan).

Tartrate-Resistant Acid Phosphatase Staining
The prepared paraffin sections were stained with a TRAP staining solution (Cosmo Bio Co., Ltd., Tokyo, Japan) according to the conventional method, and then encapsulated with a cover glass and Mount-Quick Aqueous medium. Histological images were then observed under an optical microscope. The number of TRAP-positive cells that existed within 3 mm just below the growth plate and that had at least three nuclei were counted as osteoclasts. The number of osteoclasts was measured in nine sections from three mice in each group, and three sections from each mouse were selected at random.

Immunohistochemical Staining
The prepared paraffin sections were deparaffinized, rehydrated, and stained using the avidin-biotin peroxidase system and the Vectastain Elite ABC rat kit (Vector Laboratories, Burlingame, CA, United States). Tissue sections were immersed in a 0.3% hydrogen peroxide in methanol solution for 30 min at room temperature to remove the endogenous peroxidase activity and treated with trypsin (Thermo Fisher Scientific) for 15 min for antigen activation. After blocking with normal rabbit serum for 15 min, the primary antibodies were added and reacted overnight at 4°C. For the primary antibodies, anti-cathepsin-K and antialkaline phosphatase antibodies (Abcam, Cambridge, MA, United States), which are specific marker proteins of osteoclasts and osteoblasts, were diluted (100-fold) in PBS; they were washed in PBS, and rabbit anti-rat biotin-labeled secondary antibodies were added by diluting (200-fold) in PBS and were left to react for 30 min at 25°C. The avidin-biotin-labeled enzyme complex was added and reacted for 30 min, followed by the addition of 0.01% 3,3′-diaminobenzidine (DAB; Nacalai Tesque Co., Ltd., Kyoto, Japan) for coloration. Finally, cell nuclei were contrast-stained with Mayer's hematoxylin (Wako), observed after enclosure using a cover glass and Mount-Quick, and the positive cells were counted.

Enzyme-Linked Immunosorbent Assay
Fresh blood was collected from the hearts of mice in each group. The serum was collected via centrifugation at 10,000 × g at 4°C for 15 min. The secretion of serum inflammatory cytokines (IL-1β, IL-6, TNF-α, and M-CSF) was measured using ELISA MAX Deluxe (BioLegend, San Diego, CA, United States) and Quantikine ELISA Kit (R&D systems, Minneapolis, MN, United States), following the manufacturer's instructions. The absorbance at 450 nm was measured using a microplate reader (SH-1000 Lab; Corona Electric Co., Hitachinaka, Japan). All tests were performed in two wells per individual. The serum concentrations of IL-1β, IL-6, TNF-α, and M-CSF were quantified, and the concentrations were set to 0 if the results were below the detection limit.

Mouse Bone Marrow Macrophages
Five-week-old male C57BL6/J mice were obtained from Japan CLEA (Japan). Bone marrow cells were collected from the tibiae and femur, and were cultured in M-CSF (50 ng/ml) for 3 days in culture dishes containing DMEM (Wako) supplemented with 10% fetal bovine serum (FBS, Invitrogen, Carlsbad, CA, United States). After 3 days, floating cells were removed by rinsing with phosphate-buffered saline (PBS), and the attached cells were used as mBMMs (Horibe et al., 2013;Tevlin et al., 2014). Attached mBMMs were sprayed with PBS to remove the culture dish mechanically to use for the following in vitro experiments. This animal experiment was performed in accordance with the Guidelines for Proper Conduct of Animal Experiments of the Science Councils of Japan and approved by the Animal Research Control Committee of Okayama University (approval no. OKU-2016277).

Statistical Analysis
Experimental results are presented as the mean ± standard deviation (SD). Multiple comparisons were conducted using one-way analysis of variance (ANOVA) and Tukey's test. Statistical analysis software GraphPad Prism8 (GraphPad Software Inc., San Diego, CA, United States) was used for each statistical process, and p-values lower than 0.05 were considered statistically significant.

Synthetic-(+)-Terrein-Suppressed Ovariectomy-Induced Systemic Bone Loss
OVX models were used to validate the therapeutic value of TER in suppressing osteoporosis in vivo. TER or PBS (vehicle) was intraperitoneally injected into OVX mice every 2 days for 8 weeks before euthanization. Micro CT analyses were performed to determine the effect of synthetic TER on bone mass ( Figure 3A). Compared to that of the Sham group, BMD, BV/TV, and Tb.N in the OVX + vehicle group decreased by approximately 16, 80, and 78%, respectively, whereas Tb/Sp increased by 64%. Interestingly, compared to the OVX group, after treatment with synthetic TER (30 mg/kg), BMD, BV/TV, and Tb.N increased by 7, 239, and 130%, FIGURE 5 | Synthetic TER did not affect the body weight and serum inflammatory cytokines production. (A) Effect of synthetic TER on body weight. Weight changes in Sham, OVX, OVX + TER (10-30 mg/kg, two times a week), and OVX + zoledronic acid (ZOL, 0.1 mg/kg, two times a week) groups were observed up to 2, 5, and 8 weeks after the start of the experiment as relative values (mean ± SD) of the baseline body weight value (100%) at the start of the experiment (n 4-7). The error bar: SD. (B-E) The amount of pro-inflammatory cytokines (TNF-α, IL-1β, IL-6, and M-CSF) in the serum was quantified using ELISA (n 4-7). Columns represent the mean results of the experiments carried out in, at least, triplicates, and the bars represent the SD (n 4-7, *p < 0.05, ANOVA/Tukey-Kramer test).
Frontiers in Pharmacology | www.frontiersin.org June 2021 | Volume 12 | Article 674366 respectively, whereas the relative Tb.Sp decreased by 42% ( Figure 3B, n 6-7, **p < 0.01, *p < 0.05). On the other hand, an abnormal increase in bone mass was observed in the ZOL-treated group ( Figure 3A). Histomorphometric assessments (HE staining) validated these results ( Figure 4A). Furthermore, TRAP and IHC staining were performed for qualitative and quantitative analyses of several bone parameters that reflect the effect of synthetic TER on osteoclast formation in vivo (Figures 4B,C). The number of TRAP-positive osteoclasts and cathepsin-K-positive cells decreased, together with a dramatic inhibition of osteoclast formation under treatment with synthetic TER, compared to the OVX + vehicle group ( Figures 4B-D, n 3, **p < 0.01, *p < 0.05).

Synthetic-(+)-Terrein did Not Affect Body Weight and Serum Inflammatory Cytokines Production in Ovariectomized Mice
There were no significant differences in body weight between the groups treated with/without synthetic TER (30 mg/kg) ( Figure 5A). The serum inflammatory cytokine levels of TNFα, IL-1β, IL-6, and M-CSF in the OVX mice did not differ between the groups treated with/without synthetic TER (30 mg/kg). However, in the ZOL-treated group, the secretion of inflammatory cytokines (TNF-α, IL-1β, IL-6, and M-CSF) was significantly enhanced (Figures 5B-E, n 4-7, *p < 0.05).

Synthetic-(+)-Terrein Inhibited the Phosphorylation of PKCα/βII and Receptor Activator of Nuclear Factor-kB Ligand-Induced Osteoclastogenesis in Mouse Bone Marrow Macrophages
To elucidate the potential mechanism by which TER inhibits RANKL-induced osteoclastogenesis, we focused on the PKC signaling pathway, which is also related to NFATc1 expression on RANK-RANKL signaling. RANKL stimulation strongly induced PKCα/βII protein phosphorylation at 30 min. The treatment with synthetic TER (10 μM) significantly inhibited RANKL-induced PKCα/βII protein phosphorylation, which is related to NFATc1 expression ( Figures 6A,B, n 3, *p < 0.05). On the other hand, TER did not suppress PKCδ protein phosphorylation, which is not related to NFATc1 expression ( Figures 6A,B, n 3). In addition, an osteoclastogenesis assay was performed to investigate the effect of the PKC inhibitor, GF109203X, and synthetic TER. Treatment with GF109203X resulted in the dose-dependent inhibition of RANKL-induced osteoclast formation, especially in the 10 μM GF109203X-treated group, as well as GF109203X; synthetic TER (10 μM) inhibited RANKL-induced osteoclastogenesis in mBMMs completely ( Figure 6C).

Synthetic-(+)-Terrein did Not Affect Osteoblast Differentiation in MC3T3-E1 Cells
Whether TER promotes osteoblast differentiation was also examined using mouse osteoblast-like cells, MC3T3-E1. As shown in Figure 8, TER treatment did not promote alizarin red staining and ALP staining activity, while BMP-2 enhanced both staining and activity ( Figures 8A,B, n 3; **p < 0.01). We also performed IHC staining using an ALP antibody in tissue sections extracted from OVX mice. There was no statistical difference in the number of ALP-positive cells in each group ( Figure 8C, n 3).

DISCUSSION
In this study, we demonstrated that synthetic TER is capable of suppressing RANKL-induced osteoclast differentiation by inhibiting signaling pathways (PKCα/βII) and downstream function-related gene expression (OC-stamp, DC-stamp, calcitonin receptor, and v-type protein ATPase subunit d2). An OVX mouse model further proved the therapeutic value of TER in systematic bone loss. The in vitro and in vivo results indicated that TER probably has therapeutic potential for osteolytic diseases, including osteoporosis. Bone homeostasis is maintained by the continuous progression of bone formation by osteoblasts and bone resorption by osteoclasts (Matsuo and Irie, 2008;Matsuoka et al., 2014). Overactivated osteoclasts can induce many pathological and osteochondrotic diseases, such as postmenopausal osteoporosis, rheumatoid arthritis, lytic bone metastases, Paget's disease, and periodontitis (Zaidi, 2007;Wackhoure et al., 2009;Kumar et al., 2018). In osteoporosis, estrogen deficiency leads to elevated levels of RANKL and inflammatory factors, including IL-6 and TNF-α (Lin et al., 2019). The overactivation of the RANKL signaling pathway recruits RANK/TRAF6 association, which in turn promotes osteoclast reproduction . Thus, inhibition of osteoclast formation in the RANKL signaling pathway makes it a therapeutic target for bone-destroying diseases.
In the present in vivo study, administration of synthetic TER maintained the bone mineral density, bone mass, and bone girdle count in OVX mice. Administration of ZOL, one of the BPs, enhanced the secretion of inflammatory cytokines in the serum, including TNF-α, IL-1β, IL-6, and M-CSF ( Figure 5). On the other hand, administration of synthetic TER did not enhance the secretion of inflammatory cytokines in the serum; it had a biosafety impact compared to ZOL. Therefore, synthetic TER may be a safer treatment for osteoporosis than the currently used BPs.  NFATc1 plays an indispensable role in RANKL-induced osteoclast differentiation (Asagiri and Takayanagi, 2007). NFATc1 functions downstream of NF-κB, MAPKs (ERK1/ 2, p38, and MAPK), and Ca 2+ /calcineurin signaling pathways (Miyazaki et al., 2000;Takayanagi et al., 2002;Liu et al., 2007;Negishi-Koga and Takayanagi, 2009), and its activation can stimulate the expression of bone resorption-related genes, including Atp6v0d2 (V-ATPase-d2), Itgb3 (integrin 3), Acp5 (TRAP), and Ctsk (cathepsin-K), which are essential for osteoclast differentiation (Crotti et al., 2008). Consistent with our findings, TER inhibits RANKL-induced osteoclast differentiation and bone resorption by inhibiting the expression of NFATc1 and its downstream bone resorption-related genes. On the other hand, TER did not suppress RANKL-induced phosphorylation of NF-κB, ERK, and p38 in mBMMs (Nakagawa et al., 2020). In this study, synthetic TER inhibited the RANKL-induced phosphorylation of PKCα/βII, one of the classical PKC pathways, and the PKC inhibitor, GF109203X, completely inhibited RANKL-induced osteoclastogenesis as well as synthetic TER treatment ( Figure 6C). Previous studies have found that the PKCβ pathway, leading to GSK-3β inactivation and NFATc1 induction, has a key role in the RANKL-induced osteoclast differentiation (Shin et al., 2014). In addition, the PKC inhibitor, GF109203X, suppresses RANKL-induced calcium oscillations, inhibiting calciumdependent NFAT activation in mBMMs (Yao et al., 2015). It is well known that NFATc1 levels are regulated by transcriptional control and by the actions of the Ca 2+dependent phosphatase calcineurin, which stabilizes the protein and allows its nuclear transportation and transcriptional activity (Negishi-Koga and Takayanagi, 2009). Thus, the mechanism by which TER inhibits osteoclast differentiation by suppressing PKCα/βII phosphorylation, but not NF-κB or MAPKs, was revealed, which may lead to the elucidation of a new mechanism of TER for inhibiting osteoclast differentiation via suppressing Ca 2+ oscillation and GSK-3β dynamics that require further investigation.
For the future pharmacological use of TERs, the efficacy of synthetic TERs needs to be proven in comparison to existing osteoporosis drugs, such as BPs. The long half-life of BPs is one of the problems in osteoporosis treatment for MRONJ. An excessive increase in bone hardening also conversely increases the risk of fracture. Synthetic TERs are small molecules and, therefore, may have a shorter half-life than BPs. They are also expected to provide a moderate inhibitory effect on bone resorption as a treatment for patients with early osteoporosis. Therefore, further studies in an in vivo model are needed to elucidate the systemic response of TERs.
The present study shows that synthetic TER abrogates ovariectomy-induced bone loss and RANKL-induced osteoclastogenesis by suppressing PKCα/βII phosphorylation (Figure 9). The results provide new insights into the potential use of TER as an antiresorptive agent in the treatment of osteolytic bone diseases, including osteoporosis.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
The animal study was reviewed and approved by the Animal Research Control Committee of Okayama University.

AUTHOR CONTRIBUTIONS
KS and KO contributed to the conception, design, data acquisition, analysis, and interpretation of the data, and drafted and critically revised the manuscript. MN and HM contributed to the conception, data analysis, and interpretation of the study, and critically revised the manuscript. SN, HS, CK, SY, and HK acquired, analyzed, and interpreted the data. SI (10th author) synthesized TER. MO performed micro-CT analysis. SI (12th author), KY, and TY interpreted the data. SS contributed to the conception, data analysis, and interpretation, and critically revised the manuscript. ST contributed to the conception, design, data analysis, and interpretation, and critically revised the manuscript. All authors approved the final manuscript and agreed to be accountable for all aspects of the work.

FUNDING
This study was supported by a Grant-in-Aid for Scientific Research (C) (no. 19K10108, to KO) and Young Investigator (B) (no. 20K18509, to SN) from the Japan Society for the Promotion of Science, Ryobi Memorial Foundation, and Wesco Scientific Promotion Foundation (to KO and HM).