A marine fungus‐derived nitrobenzoyl sesquiterpenoid suppresses receptor activator of NF‐κB ligand‐induced osteoclastogenesis and inflammatory bone destruction

Abstract

Background and Purpose

Osteoclasts are unique cells to absorb bone. Targeting osteoclast differentiation is a therapeutic strategy for osteolytic diseases. Natural marine products have already become important sources of new drugs. The naturally occurring nitrobenzoyl sesquiterpenoids first identified from marine fungi in 1998 are bioactive compounds with a special structure, but their pharmacological functions are largely unknown. Here, we investigated six marine fungus‐derived nitrobenzoyl sesquiterpenoids on osteoclastogenesis and elucidated the mechanisms.

Experimental Approach

Compounds were first tested by RANKL‐induced NF‐κB luciferase activity and osteoclastic TRAP assay, followed by molecular docking to characterize the structure–activity relationship. The effects and mechanisms of the most potent nitrobenzoyl sesquiterpenoid on RANKL‐induced osteoclastogenesis and bone resorption were further evaluated in vitro. Micro‐CT and histology analysis were used to assess the prevention of bone destruction by nitrobenzoyl sesquiterpenoids in vivo.

Key Results

Nitrobenzoyl sesquiterpenoid 4, with a nitrobenzoyl moiety at C‐14 and a hydroxyl group at C‐9, was the most active compound on NF‐κB activity and osteoclastogenesis. Consequently, nitrobenzoyl sesquiterpenoid 4 exhibited suppression of RANKL‐induced osteoclastogenesis and bone resorption from 0.5 μM. It blocked RANKL‐induced IκBa phosphorylation, NF‐κB p65 and RelB nuclear translocation, NFATc1 activation, reduced DC‐STAMP but not c‐Fos expression during osteoclastogenesis in vitro. Nitrobenzoyl sesquiterpenoid 4 also ameliorated LPS‐induced osteolysis in vivo.

Conclusion and Implications

These results highlighted nitrobenzoyl sesquiterpenoid 4 as a novel inhibitor of osteoclast differentiation. This marine‐derived sesquiterpenoid is a promising lead compound for the treatment of osteolytic diseases.

Abbreviations

AP‐1

activator protein

BBMs

bone marrow macrophages

DC‐STAMP

dendritic cell‐specific transmembrane protein

M‐CSF

macrophage colony‐stimulating factor

MTT

3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide

NFATc1

NF of activated T‐cell cytoplasmic 1

NS

nitrobenzoyl sesquiterpenoid

RANKL

receptor activator of nuclear factor κB ligand

TRAP

tartrate‐resistant acid phosphatase

What is already known

• Osteoclast differentiation is a target for osteolytic diseases. However safe oral osteoclast inhibitors are not available.

• Nitrobenzoyl sesquiterpenoids are bioactive and rare with few known pharmacological functions.

What does this study add

• Nitrobenzoyl sesquiterpenoid 4 decreases RANKL‐induced osteoclastogenesis in vitro and inflammatory bone loss in vivo.

• Nitrobenzoyl sesquiterpenoid 4 blocks RANKL‐induced up‐regulation of NF‐κB, NFATc1, DC‐STAMP but not c‐Fos in vitro.

What is the clinical significance

• Nitrobenzoyl sesquiterpenoid 4 could be apotential compound for treating osteolytic diseases as osteoclast differentiation inhibitor.

1. INTRODUCTION

Bone is a dynamic and hard organ that is continuously renewed and shaped. Bone metabolic homeostasis is regulated by maintaining a balance between the bone formation by osteoblasts and the bone resorption by osteoclasts (Grabowski, 2009). Excessive activation of osteoclasts interferes with this balance between osteoblasts and osteoclasts and results in the bone loss that is found in various osteolytic diseases, including osteoporosis, metastatic cancers and inflammatory disorders such as sepsis, rheumatoid arthritis and psoriatic arthritis. These diseases pose a threat to the health of hundreds millions of people around the world, although current available treatments for these diseases are not ideal (Boyce, 2013; De Vlam, Gottlieb, & Mease, 2014; Otto, Pautke, Van den Wyngaert, Niepel, & Schiødt, 2018; Tateiwa, Yoshikawa, & Kaito, 2019). In recent years, targeting osteoclast differentiation has become a feasible therapeutic strategy for preventing bone destruction, fractures and pain in osteolytic diseases (Broadhead, Clark, Dass, Choong, & Myers, 2011; Jakob et al., 2015; Panagopoulos & Lambrou, 2018). In 2012, denosumab, a receptor activator of nuclear factor κB ligand (RANKL or TNFSF11) antibody, was already one of the top 50 best‐selling drugs in the United States (Vitaku, Ilardi, & Njaroarson, 2013). However, as the representative drugs for targeting osteoclasts clinically, both the bisphosphonate zoledronic acid and denosumab have serious complications and intolerable side effects such as osteonecrosis of the jaw, which can cause disability. In addition, the other side effects of bisphosphonates include atypical femoral fractures and mouth ulcers while denosumab can also cause pain in the back and extremities, musculoskeletal pain, hypercholesterolaemia and cystitis (Otto et al., 2018; Wang, Yamauchi, & Mitsunaga, 2020). Therefore, new drugs targeting osteoclasts for treating osteolytic diseases still need to be developed.

Osteoclasts are large, unique, multinucleated cells with bone‐resorbing abilities and are shaped via the fusion of precursors derived from bone marrow mononuclear macrophage lineage (Ono & Nakashima, 2018). The formation and maturation of osteoclasts are regulated by several key factors. RANKL is essential for osteoclast differentiation and function, while macrophage colony‐stimulating factor (M‐CSF or CSF‐1) is indispensable for the survival of osteoclasts (Biskobing, Fan, & Rubin, 1995; Dougall et al., 1999). The binding of RANKL to its cognate receptor RANK triggers a cascade of intracellular signalling events, including the activation of NF‐κB, activator protein (AP‐1) and nuclear factor of activated T‐cell cytoplasmic 1 (NFATc1) transcription factors (Boyle, Simonet, & Lacey, 2003; Hirotani, Tuohy, Woo, Stern, & Clipstone, 2004; Matsuo et al., 2004). NF‐κB plays an important role in the osteoclast differentiation and function (Boyce, Xing, Franzoso, & Siebenlist, 1999; Ghosh & Karin, ; Jimi & Ghosh, 2005), and NF‐κB can be activated through classical and alternative pathways (Jimi & Ghosh, 2005). Activated NF‐κB can also lead to the induction of c‐Fos (subunit of AP‐1) and NFATc1 (Jimi & Ghosh, 2005; Takatsuna et al., 2005). Finally, NFATc1 can auto‐amplify and cooperate with other nuclear factors to induce osteoclast‐related genes and fusion genes to initiate osteoclastogenesis (Hirotani et al., 2004; Matsuo et al., 2004).

Sesquiterpenoids, primarily found in higher plants, marine organisms and microorganisms, have provided researchers with many encouraging leads for chemotherapeutic agents and treatments of immune diseases (Durán‐Peña, Ares, Hanson, Collado, & Hernández‐Galán, 2015; Qin, Wang, & Zhang, 2017). Marine organisms have become important sources of new drugs (Kang, Seo, & Park, 2015; Malve, 2016). Marine‐derived sesquiterpenoids exhibit a wide range of biological activities, including antimicrobial activity, anti‐inflammatory activity and anticancer activity (Cheung, Ng, Wong, Chen, & Chan, 2016; El‐Kassem, Hawas, El‐Desouky, & Al‐Farawati, 2018). Nitrobenzoyl sesquiterpenoids (NSs) are naturally rare. To date, only seven nitrobenzoyl sesquiterpenoids have been reported and among these, three were discovered and reported by us, namely, 6β, 9α‐dihydroxy‐14‐p‐nitrobenzoylcinnamolide (NS4), insulicolide B (NS1) and insulicolide C (NS3) from the marine‐derived fungus Aspergillus ochraceus Jcma1F17. Although these nitrobenzoyl sesquiterpenoids have a special structure and exhibit pharmacological activities, only their simple cytotoxic effects on tumour cells along with their anti‐viral and anti‐inflammation activity in vitro have been addressed (Belofsky, Jensen, Renner, & Fenical, 1998; Fang et al., 2014; Rahbaek et al., 1997; Tan et al., 2018; Wu et al., 2012; Zhao et al., 2016). Despite the above, marine‐derived nitrobenzoyl sesquiterpenoids have already attracted scientific attention. For instance, the total chemical synthesis of 4 nitrobenzoyl sesquiterpenoids has recently been published, thereby providing more sources of these uncommon nitrobenzoyl sesquiterpenoids for medicinal research and other future applications (Lai, Zhang, Zhang, Chen, & Yang, 2018). A deeper understanding of the pharmacological functions of nitrobenzoyl sesquiterpenoids is now warranted.

Here, we have isolated nitrobenzoyl sesquiterpenoids and evaluated the inhibitory activities of nitrobenzoyl sesquiterpenoid 1–6 and 7 by RANKL‐induced NF‐κB luciferase activity and osteoclastic TRAP assays in pre‐osteoclastic RAW264.7 cells. The compound nitrobenzoyl sesquiterpenoid 4 was selected for further study of its effects and underlying mechanisms in osteoclast formation and absorption in vitro and on bone destruction in vivo. Our research discovered that marine fungus‐derived nitrobenzoyl sesquiterpenoid 4 could be a promising osteoclast differentiation inhibitor, which could have therapeutic benefits for osteoclast‐related osteolytic diseases.

2. METHODS

2.1. Reagents and antibodies

The extraction process and structures of nitrobenzoyl sesquiterpenoids were described in our previous study (Tan et al., 2018). RAW264.7 cells (RRID:CVCL_0493) were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). NF‐κB‐Luc and NFATc1‐Luc transfected RAW264.7 cells (Zhou et al., 2016) were kindly provided by Prof. Jiake Xu (University of Western Australia, Nedlands, Australia). DMEM, α modification of Eagle's medium (α‐MEM), and FBS were obtained from Gibco (Rockville, MD, USA). Recombinant mouse RANKL and M‐CSF were purchased from R&D Systems (Minneapolis, MN, USA). MTT, TRAP Kit, cellcounting kit 8, BAY11‐7082, propidium iodide (PI), toluidine blue, calcein, DAPI and CsA were provided by Sigma‐Aldrich (St Louis, MO, USA). Lactate dehydrogenase activity assay kit was bought from Jiancheng Bioengineering Institute (Nanjing, China). A RNeasy mini kit (Qiagen, Valencia, CA, USA) and a PrimeScript RT reagent kit were obtained from TaKaRa (Dalian, China). Luciferase reagents and GoTaq® qPCR Master Mix were provided by Promega (Madison, WI, USA). Hydroxyapatite‐coated plates were provided by Corning Life Science (St. Lowell, MA, USA) and bone slices were provided by IDS (London, UK). A nuclear extraction kit was provided by Cayman Chemicals (Ann Arbor, MI, USA). Alexa Fluor 488 conjugated secondary antibody, Opti‐MEM, and Lipofectamine™ 2000 were purchased from Invitrogen (Carlsbad, USA). Anti‐DC‐STAMP rabbit Ab was purchased from OmnimAbs (OM203458). Rabbit mAbs specific to p‐IκBa (#2859, RRID:AB_561111), NF‐κB p65 (#49445, RRID:AB_2799359), RelB (#10544, RRID:AB_2797727), NFATc1 (#8032, RRID:AB_10829466), c‐Fos (#2250, RRID:AB_2247211) and β‐actin (#3700, RRID:AB_2242334) and mouse mAb specific to lamin A/C (#4777, RRID:AB_10545756) were all from Cell Signalling Technology (Beverly, MA, USA).

2.2. Mice

Eight‐ to 12‐week‐old female ICR mice (25–30 g, RRID:MGI:5652673) and C57/BL6 mice (18–22 g, RRID:MGI:5658887) were purchased from the medical animal centre of Guangdong province, China. The mice were housed in micro‐isolator cages under conditions at 22–24°C with a 12‐h light/dark cycle. Water and food were provided ad libitum. The experimental protocols were approved by the Institutional Animal Care and Use Committee of the Southern Medical University. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny et al., 2010) and with the recommendations made by the British Journal of Pharmacology.

2.3. Cell viability assay

MTT and cellcounting kit 8 were used to evaluate the cytotoxic effects of nitrobenzoyl sesquiterpenoids on RAW264.7 cells and nitrobenzoyl sesquiterpenoid 4 on bone marrow macrophages (BMMs), respectively. RAW264.7 cells (3 × 10⁵ cells·ml⁻¹) were cultured with or without nitrobenzoyl sesquiterpenoids (4 μM) for 12 h. RAW264.7 cells (1 × 10³ cells·ml⁻¹) were cultured with or without nitrobenzoyl sesquiterpenoids (0.5, 1, and 2 μM) for 5 days. BMMs (1 × 10⁵ cells·ml⁻¹) with 50 ng·ml⁻¹ M‐CSF were seeded with or without nitrobenzoyl sesquiterpenoid 4 (0.5, 1 or 2 μM) for 24 h or 5 days. Cell viability was assessed according to manufacturer's instructions; the results are carried out as ratio to the control.

2.4. Osteoclast differentiation and TRAP assay

Pre‐osteoclastic RAW264.7 cells cultured with the compounds (nitrobenzoyl sesquiterpenoid 1–6 or 7) at 1 μM were activated by RANKL (100 ng·ml⁻¹) for 5 days to screen the most active compound on osteoclastogenesis. For further study of nitrobenzoyl sesquiterpenoid 4 on osteoclastogenesis, nitrobenzoyl sesquiterpenoid 4 at 0.5 to 2 μM concentrations were added in RAW264.7 cells with RANKL stimulation for 5 days. Meanwhile, BMMs were extracted from C57/BL6 mice as previously described (Guan, Zhao, Cao, Chen, & Xiao, 2015). BMMs (1 × 10⁵ cells·ml⁻¹) were incubated in a 96‐well plate with nitrobenzoyl sesquiterpenoid 4 (0.5, 1 or 2 μM), followed by stimulation with RANKL (100 ng·ml⁻¹) plus M‐CSF (50 ng·ml⁻¹) for approximately 3 days for osteoclast differentiation. Then the cells were fixed and stained for tartrate‐resistant acidic phosphatase activity (TRAP) and the images were photographed by using an inverted microscope (Nikon, Japan). TRAP‐positive multinucleated cells, nuclei >3 for RAW264.7 cells and nuclei >5 for BMMs, were regarded as osteoclasts (Ouyang et al., 2014; Wu et al., 2012).

2.5. Bone resorption pit assay

RAW264.7 cells (1 × 10³ cells·ml⁻¹) and BMMs (1 × 10⁴ cells·ml⁻¹) were seeded in hydroxyapatite‐coated plates, administered with nitrobenzoyl sesquiterpenoid 4 at different doses (0.5, 1 or 2 μM) and then stimulated by RANKL (100 ng·ml⁻¹) for 7 days or RANKL (100 ng·ml⁻¹) plus M‐CSF (50 ng·ml⁻¹) for 5 days, respectively. The supernatants were collected from BMMs to measure the release of Ca⁺⁺ from the plates by using the calcium detection kit (BestBio, China); the plates were washed with 10% bleach solution for pit assay. Furthermore, BMMs (1 × 10⁴ cells·ml⁻¹) were plated on bone slices in 96‐well plates with nitrobenzoyl sesquiterpenoid 4 (2 μM), followed by stimulation with RANKL (100 ng·ml⁻¹) plus M‐CSF (50 ng·ml⁻¹). Seven days later, cells were removed and stained by 1% toluidine blue. Osteoclast resorbing pits were then documented by a light microscope (IX71; Olympus) and Image‐Pro Plus 6 software (RRID:SCR_007369) was employed to measure the pit areas.

2.6. NF‐κB luciferase assay

The effects of nitrobenzoyl sesquiterpenoid 1–6 and 7 on RANKL‐induced NF‐κB luciferase activity were detected by luciferase reporter gene assay as described earlier. Briefly, RAW264.7 cells, stably transfected with a NF‐κB luciferase reporter construct, were treated with nitrobenzoyl sesquiterpenoid 1–6 or 7 (4 μM) and BAY11‐7082 (NF‐κB inhibitor, 5 μM) for 4 h, followed by stimulation with RANKL (100 ng·ml⁻¹) for 6 h. The luciferase activity was determined using the luciferase assay system (Promega, Madison, WI). The dose‐dependent effects of nitrobenzoyl sesquiterpenoid 4 (0.5–2 μM) on RANKL‐induced NF‐κB luciferase activity were detected by the same assay.

2.7. NF‐κB nuclear translocation assay by confocal microscope

RAW264.7 cells (2 × 10⁵ cells·ml⁻¹) were seeded in confocal plates overnight, followed by adding serum‐free medium for another 24 h. Then cells were treated with nitrobenzoyl sesquiterpenoid 4 (2 μM) for 4 h, further stimulated with 100 ng·ml⁻¹ RANKL for 30 min, fixed using 4% paraformaldehyde for 15 min, permeabilized by 0.1% Triton‐X 100 for 30 min and blocked with 10% goat serum for 60 min at room temperature. The cells were incubated with the NF‐κB p65 antibody (1:200) in 10% goat serum solution overnight at 4°C and then stained with Alexa Fluor 488 conjugated secondary antibody for 1 h and incubated with DAPI solution (1 μg·ml⁻¹) for 5 min in the dark for nuclear staining. After the cells were washed with PBS, the nuclear translocation of NF‐κB p65 was imaged by using an inverted laser confocal microscope (LSM 880 with Airyscan, Zeiss).

2.8. Molecular docking of nitrobenzoyl sesquiterpenoids with NF‐κB p65 and nitrobenzoyl sesquiterpenoid 4 with IκBa

The docking studies were performed using molecular operating environment, 2016 software (Chemical Computing Group, Montreal, Canada). The crystal structure of NF‐κB p65 with 1.8 Å resolution (PDB: 1MY5) was obtained from the protein data bank (http://www.rcsb.org/pdb/home/home.do). Homology modelling of IκBa was used because there is no crystal structure in the PDB database. By searching the UniProt database (RRID:SCR_004426) with UniProt ID of IκBa: Q9Z1E3, we found that there were two templates PDB: 1IKN and 1NFI with similarity of 93–94%. We downloaded the template 1NFI from SWISS MODEL (https://swissmodel.expasy.org/repository, RRID:SCR_013032). According to the requirement for the docking study, water molecules, ions, and nonstandard amino acid residues were detached from the proteins, and the hydrogen atoms were added under an AMBERT10 force field. After the automated correction of protein structure using the “Structure Preparation” module, the binding sites were selected using the “Site Finder” procedure. The prepared ligand was then flexibly docked into the receptor using the “Triangle Matcher” placement method and “London dG” scoring with other default parameters. Finally, five docking poses were obtained, and the one with the best score was chosen for the structure of nitrobenzoyl sesquiterpenoids.

2.9. Plasmids construct and transfection

The overexpression vectors carried murine WT and mutant NF‐κB p65 and IκBa plasmids were constructed by Kidan Bio Co. Ltd (Guangzhou, China). Briefly, WT NF‐κB p65 and mutant NF‐κB p65 (R246A), WT IκBa and mutant IκBa (K132A) were amplified by PCR with the primers (with XbaI and NotI restriction sites, respectively) and then the amplified fragments were digested with XbaI and NotI, cloned into vector pCDH‐CMV‐MCS‐EF1‐CopGFP‐T2A‐Puro. Arginine (Arg) was mutated to alanine (Ala) at site 246 of NF‐κB p65 in mutated NF‐κB p65 (R246A) plasmid, and lysine (Lys) was mutated to Ala at site of 132 of IκBa protein in mutant IκBa (K132A) plasmid. The constructed plasmids were finally sequenced to confirm the site mutation.

The mutant or control plasmids (a total of 1 μg) mixed with Lipofectamine™ 2000 at a ratio of 1:1 in a total volume of 100μl Opti‐MEM, were firstly added to RAW264.7 cells or RAW264.7 cells stably transfected with a NFATc1 luciferase reporter construct (NFATc1‐Luc‐RAW264.7 cells) for 6 h. The supernatants were replaced by fresh DMEM medium and the transfected cells were allowed to grow for 42 h for further stimulation or transfection assay. GFP was observed as an indicator to identify transfection‐positive cells by fluorescence microscope and western blot analysis was done to verify if NF‐κB p65 and IκBa protein were overexpressed after WT, mutant plasmids transfection compared with empty vector.

2.10. Quantitative real‐time PCR assay

For real‐time PCR, RAW264.7 cells (3 × 10⁵ cells ml⁻¹) were cultured with nitrobenzoyl sesquiterpenoid 4 at different concentrations (0.5, 1 or 2 μM) for 4 h and then stimulated with RANKL (100 ng·ml⁻¹) for 24 h. Total RNA was extracted using the RNeasy mini kit, cDNA was synthesized using the PrimeScript RT reagent kit, and real‐time PCR was executed using qPCR Master Mix as previously described. The primer sequences are listed in Table 1. The relative expression of analysed genes was normalized to internal control GAPDH.

TABLE 1.

The sequences of gene primers

Gene name	Primer sequence
TRAP R	ACACAGTGATGCTGTGTGGCAACTC
TRAP F	CCAGAGGCTTCCACATATATGATGG
β3‐integrin R	TGACATCGAGCAGGTGAAAG
β3‐integrin F	GAGTAGCAAGGCCAATGAGC
DC‐STAMP R	TCCTCCATGAACAAACAGTTCCAA
DC‐STAMP F	AGACGTGGTTTAGGAATGCAGCTC
NFATc1 R	GGAAGTCAGAAGTGGGTGGA
NFATc1 F	GGGTCAGTGTGA CCGAAGAT
GAPDH R	AACTTTGGCATTGTGGAAGG
GAPDH F	ACACATTGGGGGTAGGAACA

2.11. Western blot analysis

RAW264.7 cells were first cultured with different doses of nitrobenzoyl sesquiterpenoid 4 for 4 h before stimulation. After the incubation with RANKL (100 ng·ml⁻¹) for 30 min (for detecting p65, p‐IκBa, RelB, 1 × 10⁶ cells·ml⁻¹), 24 h (for c‐Fos, NFATc1 detection, 3 × 10⁵ cells·ml⁻¹) or 48 h (for DC‐STAMP detection, 1 × 10⁵ cells·ml⁻¹) whole cell lysates were prepared using RIPA buffer while cytoplasmic extracts and nuclear extracts were prepared using a nuclear extraction kit according to the manufacturer's protocol. Protein concentrations of cells, cytoplasmic and nuclear extracts were measured by bicinchoninic acid assay. Proteins were electrophoresed on 10% SDS‐PAGE and transferred to PVDF membranes (Millipore, MA, USA). The membranes were blocked using 5% non‐fat milk in TBST for 1 h at 37°C, followed by incubation with the primary antibodies overnight at 4°C. Next, the membranes were washed with TBST and incubated for 1 h with horseradish peroxidase‐conjugated secondary antibody at 37°C and finally detected using the enhanced chemiluminescent system (Yeasen Biotech, Shanghai, China) in a multifunctional imaging analysis system (Protein Simple, California, USA). Quantification of protein was detected by using ImageJ software (RRID:SCR_003070) and the data were normalized to β‐actin or lamin A/C.

RAW264.7 cells, transfected with empty vector control, WT plasmids, or site‐mutant NF‐κB p65 R246A or IκBa K132A plasmids, were incubated with or without RANKL (100 ng·ml⁻¹) for 30 min and then the total proteins, cytoplasmic extracts and nuclear extracts were prepared as described above to detect p65, IκBa, p65 nuclear translocation and p‐IκBa respectively. The immuno‐related procedures used comply with the recommendations made by the British Journal of Pharmacology (Alexander et al., 2018).

2.12. NFATc1 luciferase assay

RAW264.7 cells, stably transfected with a NFATc1 luciferase reporter construct (NFATc1‐Luc‐RAW264.7 cells), were pretreated with the nitrobenzoyl sesquiterpenoid 4 (0.5–4 μM) and cyclosporin A (1 μM) (NFATc1 inhibitor) for 4 h, followed by stimulation with RANKL (100 ng·ml⁻¹) for 24 h. The luciferase activity was determined using the luciferase assay system (Promega, Madison, WI, USA).

To value the effect of NF‐κB p65 Arg B246 mutation on NF‐ATc1 luciferase activity during osteoclastogenesis in vitro, the cells above were transfected with NF‐κB p65 R246A or control plasmids first, pretreated with the nitrobenzoyl sesquiterpenoid 4 (2 μM) for 4 h and then the NFATc1 luciferase activity was determined after RANKL stimulation for 24 h.

2.13. In vivo murine model of LPS‐induced inflammatory osteolysis

Female ICR mice (8–12 weeks old) were randomly divided into four groups (n = 6 per group) and adhered to researcher‐blinded analyses: Control, LPS, LPS + nitrobenzoyl sesquiterpenoid 4 (1 mg·kg⁻¹) and LPS + nitrobenzoyl sesquiterpenoid 4 (5 mg·kg⁻¹) groups. LPS (5 mg·kg⁻¹ body weight) was injected intraperitoneally on the first and fourth days to induce inflammatory bone loss (Zhai et al., 2014). Nitrobenzoyl sesquiterpenoid 4 and PBS controls were orally administered daily via gavage 1 day before the injection of LPS till the end of the experiment. Eight days after nitrobenzoyl sesquiterpenoid 4 treatment mice were killed by cervical dislocation and their left femurs were removed and scanned by a high‐resolution microcomputed tomography (micro‐CT) (μCT80, ScancoMedical, Zurich, Switzerland) with the following instrument parameters:‐ 50 kV, 500 μA, 0.7° rotation step. After reconstruction, a region of interest (0.5 mm above the growth plate on distal femur with a height of 1 mm) was selected for further qualitative and quantitative. The parameters of trabecular bone, including BMD (bone mineral density), ConnD (connectivity density), BV/TV (bone volume/tissue volume), Tb. N (trabecular number), Tb. Th (trabecular thickness) and Tb. Sp (trabecular separation) were measured. In addition, cortical bone was also analysed by measuring the average thickness of both cortices (Cor.Th). The right femurs of all animals were collected and fixed in 4% paraformaldehyde at 4°C for 1 day, decalcified in 12% EDTA for a month and then embedded in paraffin. Sequential sections were prepared for HE or TRAP assay. As for HE and TRAP staining images, there were n = 12 images taken in total per group (two images from each mouse). A 600 μm × 600 μm region of interest located 150 μm below the growth plate of the femur metaphysis was employed for the assessment of the trabecular bone volume (BV/TV), number of TRAP positive multinucleated cells and osteoclastic surface/bone surface (Oc.S/BS). Histopathological analysis was performed using Image Pro Plus 6.0 (IPP) software.

Another female ICR mice (8–12 weeks old) were randomly divided into three groups (n = 5 per group): Control, LPS and LPS + nitrobenzoyl sesquiterpenoid 4 (5 mg·kg⁻¹) groups as described above for calcein labelling assay. All the mice were injected intraperitoneally with calcein (20 mg·kg⁻¹ body weight in 2% NaHCO₃) at Days 1 and 5 (Conesa‐Buendía et al., 2019). At Day 8 mice were killed, their left femurs were dehydrated in gradient ethanol (60%, 70%, 80%, 90% and 100%) and then embedded in polymethyl methacrylate resin for undecalcificated bones slicing. Longitudinal sections were observed using inverted fluorescence microscope. All the above steps are required to avoid light. The average distances between fluorescent bands were measured by IPP software. MAR, one of the major dynamic parameters to fully evaluate the bone formation, was calculated (Dempster et al., 2013). MAR (μm/day) = (average distances between fluorescent bands)/(number of days between calcein injections).

2.14. Data analysis

The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology (Curtis et al., 2018). Sample sizes subjected to statistical analysis at least n = 5, where n = number of independent values. Data are expressed as the mean ± SD, one‐way ANOVA followed by Dunnett's post hoc test was used only if F achieved P values < 0.05 and there was no significant variance in homogeneity. The P values <0.05 were regarded as statistically significant.

3. RESULTS

3.1. Effects of nitrobenzoyl sesquiterpenoids on RANKL‐induced NF‐κB luciferase activity and osteoclastogenesis in pre‐osteoclastic RAW264.7 cells

Six nitrobenzoyl sesquiterpenoids, namely, insulicolide B (NS1), 14‐O‐acetylinsulicolide A (NS2), insulicolide C (NS3), 6β, 9α‐dihydroxy‐14‐p‐nitrobenzoylcinnamolide (NS4), insulicolide A (NS5), 9‐deoxyinsulicolide A (NS6), and a derivative sesquiterpenoid (7) (Figure 1a), were obtained from the marine‐derived fungus Aspergillus ochraceus Jcma1F17, as reported earlier (Tan et al., 2018).

FIGURE 1.

Inhibition of RANKL‐induced NF‐κB luciferase activity and osteoclastogenesis by nitrobenzoyl sesquiterpenoids (NS1–6 and 7), respectively. Structures of NS1–6 and 7 (a). RAW264.7 cells which had been transfected with an NF‐κB luciferase reporter construct were cultured with NS1–6, 7 (4 μM) or NF‐κB inhibitor BAY11‐7082 (5 μM) for 4 h, and then stimulated by RANKL (100 ng·ml⁻¹). Six hours later the luciferase activity (b) was measured. Cell viability of NS1–6 or 7 (4 μM) for 12 h in RAW264.7 cells (c) were measured by MTT assay. RAW264.7 cells treated with NS1–6 or 7 (1 μM) and RANKL for 5 days, TRAP‐positive multinucleated cells (nuclei >3, purple) were regarded as osteoclasts, some osteoclasts were indicated by red arrows (d). Quantification of osteoclasts treated with NS1–6 and 7 were shown (e). Cell viability of NS1–6 or 7 (1 μM) for 5 days (f) on RAW264.7 cells were measured by MTT assay. Values are expressed as the means ± SD (n = 5 independent experiments). ^# P < 0.05 relative to untreated controls, *P < 0.05 compared to RANKL‐treated controls. Scale bars, 500 μm

Because the NF‐κB pathway is one of the major signalling pathways in osteoclast differentiation and activated by RANKL (Abu‐Amer, Darwech, & Otero, 2008; Jimi & Ghosh, 2005), the inhibitory effects of the six nitrobenzoyl sesquiterpenoids (NS1–6) and one derivative sesquiterpenoid (7) on RANKL‐up‐regulated NF‐κB luciferase activity were evaluated in NF‐κB‐Luc stably transfected RAW264.7 cells. The compounds nitrobenzoyl sesquiterpenoids 4 and nitrobenzoyl sesquiterpenoids 1 at 4 μM concentration showed inhibition (34.7% and 15.5%, respectively) on NF‐κB luciferase activity induced by RANKL. None of the other compounds had any obvious inhibition of RANKL‐increased NF‐κB luciferase activity (Figure 1b).

Next, the effects of nitrobenzoyl sesquiterpenoids 1–6 and 7 on osteoclast differentiation were assessed using TRAP assays. Interestingly, 1 μM nitrobenzoyl sesquiterpenoids 4 significantly reduced osteoclast formation induced by RANKL in pre‐osteoclastic RAW264.7 cells with an inhibitory rate of 52.1% (Figures 1d,e). At the same time, none of the other compounds inhibited osteoclast formation at 1 μM obviously. None of the compounds had cytotoxic effects on RAW264.7 cells when treated for 12 h at 4 μM and 5 days at 0.5, 1 and 2 μM (Figures 1c,f and S1). Taken together, nitrobenzoyl sesquiterpenoids 4 was the most active compound tested to suppress RANKL‐induced NF‐κB luciferase activity and osteoclast formation. Therefore, it was first supposed that nitrobenzoyl sesquiterpenoid 4 could attenuate osteoclastogenesis by affecting NF‐κB.

3.2. Among the nitrobenzoyl sesquiterpenoids, nitrobenzoyl sesquiterpenoid 4 binds well to NF‐κB p65 by molecular docking

To further evaluate NF‐κB as an important molecular target for nitrobenzoyl sesquiterpenoids to inhibit osteoclastogenesis, we performed a docking study to analyse the possible binding of nitrobenzoyl sesquiterpenoids and NF‐κB p65 protein. From the proposed binding mode, we could see that the stereo‐conformation of nitrobenzoyl sesquiterpenoid 4 fit with the binding site around the reactive residue, Arg B246 (Figure 2a). The hydroxyl at C‐6 and the carbonyl group of nitrobenzoyl at C‐14 in the structure of nitrobenzoyl sesquiterpenoid 4 form additional hydrogen bonds with the residue Arg B246, whereas the hydroxyl at C‐9 forms a hydrogen bond with Lys B221 (Figure 2b). Docking studies of all other compounds tested (nitrobenzoyl sesquiterpenoid 1–3, 5–6 and 7) on NF‐κB p65 are presented in Figure S2. Nitrobenzoyl sesquiterpenoid 4 showed the strongest binding to the NF‐κB p65 protein after analysis. The hydroxyl groups at C‐6 and C‐9 in the structures of the nitrobenzoyl sesquiterpenoids were suggested to be effective groups for binding to the p65 protein. The binding activities of nitrobenzoyl sesquiterpenoids were also in line with their effects in the osteoclastic TRAP assay.

FIGURE 2.

Molecular docking of nitrobenzoyl sesquiterpenoid 4 with NF‐κB p65. (a) Binding sites of the molecule nitrobenzoyl sesquiterpenoid 4 (NS4) with the NF‐κB p65 protein. Hydrophobic, polar and the exposed regions of the receptor are depicted in green, purple, and red colours, respectively. (b) The interaction details of the predicted binding mode of NS4 with p65. The contact residues are displayed and labelled by type and number, along with detailed interaction types, distance and energy

3.3. Nitrobenzoyl sesquiterpenoid 4 suppresses the osteoclast formation and bone resorption induced by RANKL in vitro

Based on the results above, we focused on the effects of nitrobenzoyl sesquiterpenoid 4 during osteoclastogenesis. First, RAW264.7 cells and BMMs as precursor cells were cultured with various doses of nitrobenzoyl sesquiterpenoid 4 (0.5, 1 or 2 μM) and then stimulated with RANKL or RANKL + M‐CSF, respectively. The results showed that nitrobenzoyl sesquiterpenoid 4 obviously suppressed osteoclast formation at 0.5 and 2 μM in the TRAP assay by pre‐osteoclastic RAW264.7 cells (Figure 3a). Additionally, nitrobenzoyl sesquiterpenoid 4 at concentrations of 0.5–2 μM inhibited osteoclast formation in BMMs (Figure 3b). Nitrobenzoyl sesquiterpenoid 4 significantly attenuated 80.4% of osteoclast formation at 2 μM, and its inhibitory activity was found to be dose‐dependent (Figures 3b). This activity was similar to its effects on RAW264.7 cells (Figures 3a); 0.5 to 2 μM nitrobenzoyl sesquiterpenoid 4 showed no cytotoxic effect on RAW264.7 cells and BMMs by MTT or cellcounting kit 8 assays (Figure 3g,h,i). In addition, lactate dehydrogenase assay and PI staining analysed by flow cytometry to detect dead cells also indicated that there were no obvious toxicity to osteoclastic precursor cells by nitrobenzoyl sesquiterpenoid 4 (Figure S3). Many chemicals can prompt osteoblast to increase bone mass, in this respect 2 μM nitrobenzoyl sesquiterpenoid 4 did not affect the proliferation of osteoblast cells (Figure S4).

FIGURE 3.

Nitrobenzoyl sesquiterpenoid 4 (NS4) suppresses RANKL‐induced osteoclast differentiation and functions. Representative images of osteoclasts from RAW264.7 cells treated with NS4 (0.5–2 μM) for 5 days, TRAP‐positive multinucleated cells (nuclei >3, purple) were regarded as osteoclasts and indicated by red arrows (a, left) and quantified (a, right). BMMs were cultured with M‐CSF (50 ng·ml⁻¹) and NS4 (0.5, 1 or 2 μM), then stimulated by 100 ng·ml⁻¹ of RANKL for 3 days. TRAP‐positive multinucleated cells (nuclei >5) were regarded as osteoclasts (b, left) and quantified (b, right). Representative images of bone resorption area on the hydroxyapatite‐coated surfaces by osteoclasts from pre‐osteoclastic RAW264.7 cells (c) or BMMs (d) are shown. The release of Ca⁺⁺ from the hydroxyapatite‐coated plates to the culture medium after 2 μM NS4 administration were also measured (e). The resorption by osteoclasts on bone slice was also inhibited by 2‐μM NS4 (f). The resorption area by osteoclasts from RAW264.7 cells (c, right) or BMMs (d, f, right) was quantified as a percentage to the total area of hydroxyapatite‐coated surface or bone slice surface, and some resorption areas are indicated by red arrows. Cell viability of NS4 (g) at 0.5–2 μM in RAW264.7 cells for 5 days were measured by MTT assay. Cell viability of NS4 at different concentrations on BMMs for 24 h (h) and 5 days (i) were measured by cellcounting kit 8 assay. Values are expressed as the means ± SD (n = 5 independent experiments). ^# P < 0.05 relative to untreated controls, *P < 0.05 relative to RANKL‐treated controls. Scale bars, 500 μm

As a next step, the resorptive function of nitrobenzoyl sesquiterpenoid 4 on osteoclasts was evaluated in RAW264.7 cells and BMMs in hydroxyapatite‐coated plates. Nitrobenzoyl sesquiterpenoid 4 administration resulted in a restrained bone resorption area in a dose‐dependent manner in both RAW264.7 cells and BMMs after RANKL stimulation (Figure 3c,d). The resorption areas caused by osteoclasts after RANKL activation were significantly decreased by 0.5, 1 and 2 μM nitrobenzoyl sesquiterpenoid 4 in RAW264.7 cells causing an inhibition of 16.09%, 45.98% and 62.64% (Figure 3c), while in BMMs the inhibition was 9.87%, 42.86% and 73.47%, respectively (Figure 3d). Additionally, we found nitrobenzoyl sesquiterpenoid 4 (2 μM) decreased the release of Ca⁺⁺ from the hydroxyapatite‐coated plates to the culture medium (Figure 3e). Interestingly, we also observed that nitrobenzoyl sesquiterpenoid 4 (2 μM) inhibited resorption by osteoclasts in BMMs after RANKL stimulation on bone slices (Figure 3f). These results further demonstrate that nitrobenzoyl sesquiterpenoid 4 could reduce both RANKL‐induced osteoclast differentiation and bone resorptive function beginning at 500 nM in vitro.

3.4. Nitrobenzoyl sesquiterpenoid 4 inhibits the RANKL‐induced NF‐κB signalling pathway

As NF‐κB was identified as a target of bioactive nitrobenzoyl sesquiterpenoids on osteoclastogenesis by NF‐κB luciferase activity and protein docking analysis, we then further elucidated the effect of nitrobenzoyl sesquiterpenoid 4 on the NF‐κB signalling pathway. Here, we first observed that nitrobenzoyl sesquiterpenoid 4 showed dose‐dependent inhibition on RANKL‐induced NF‐κB luciferase activity (Figure 4a), which indicated nitrobenzoyl sesquiterpenoid 4 could inhibit the transcription and expression of NF‐κB.

FIGURE 4.

Nitrobenzoyl sesquiterpenoid 4 (NS4) suppresses NF‐κB signalling pathway induced by RANKL. RAW264.7 cells stably transfected with an NF‐κB luciferase reporter construct were cultured with NS4 (0.5, 1, and 2 μM) for 4 h and then stimulated by RANKL, 6 h later the luciferase activity (a) was measured. RAW264.7 cells were cultured with NS4 (0.5, 1, or 2 μM) for 4 h, and then stimulated by RANKL (100 ng·ml⁻¹) for 30 min. Total proteins, cytosolic, and nuclear proteins were extracted and analysed by western blotting using antibodies against p‐IκBa, β‐actin, p65, RelB, and lamin A/C. The relative protein or nuclear protein expression levels of p‐IκBa (b, c) and p65 (d, e) to β‐actin, or p65 (f, g) and RelB (i, j) to lamin A/C were determined using ImageJ software. RAW264.7 cells were cultured with NS4 (2 μM) for 4 h and then stimulated by RANKL (100 ng·ml⁻¹) for 30 min. The nuclear translocation of NF‐κB p65 (h) was imaged by immunofluorescence analysis. Values are expressed as the means ± SD (n = 5 independent experiments). ^# P < 0.05, relative to untreated controls, *P < 0.05 relative to RANKL‐treated controls. Scale bars, 10 μm

NF‐κB can be activated through classical and alternative pathways after RANKL stimulation during osteoclastogenesis (Abu‐Amer, 2013). In the classical pathway, IκBa phosphorylates and degrades followed by the release of p65 to translocate into nucleus, where p65 binds to the promoter regions of some osteoclastic genes to prompt their expressions (Boyce et al., 1999; Jimi & Ghosh, 2005; Abu‐Amer, 2013). Nitrobenzoyl sesquiterpenoid 4 markedly suppressed RANKL‐induced IκBa phosphorylation (Figure 4b,c). In molecular docking analysis, nitrobenzoyl sesquiterpenoid 4 showed direct binding to IκBa protein at residue Gly99 and Lys 132, and Lys 132 is the strongest binding site of IκBa with nitrobenzoyl sesquiterpenoid 4 (Figure S5A, B). While there is no report yet that Gly99 and Lys132 are sites where phosphorylation occurs in IκBa. Here, we constructed a plasmid IκBa K132A which carries the mutation of Lys to Ala at site 132 of IκBa. WT and mutant IκBa (K132A) plasmids were successfully transfected to RAW264.7 cells compared with empty vector (Figure S5C, D). After RANKL stimulation during osteoclastogenesis in pre‐osteoclastic RAW264.7 cells, the level of pIκBa was not affected after mutated IκBa K132A plasmids transfection compared with that of WT control (Figure S5E). It showed that IκBa K132A mutation did not affect the pIκBa after RANKL stimulation. This indicated that nitrobenzoyl sesquiterpenoid 4 could not reduce the IκBa phosphorylation directly by binding to Lys132 of IκBa protein. Nitrobenzoyl sesquiterpenoid 4 might inhibit IκBa phosphorylation by disturbing the structural change of whole IκBa protein during its phosphorylation or by other mechanism. Next, western blot analysis showed that nitrobenzoyl sesquiterpenoid 4 treatment significantly increased NF‐κB p65 protein levels in cytoplasm, while it inhibited NF‐κB p65 nuclear protein levels after RANKL stimulation in RAW264.7 cells in a dose‐dependent manner (Figure 4d–g). In addition, the effect of nitrobenzoyl sesquiterpenoid 4 on NF‐κB nuclear translocation was also confirmed by immunofluorescence assays. As shown in Figure 4h, p65 localized in the cytoplasm of RAW264.7 cells without RANKL activation. After stimulation with RANKL for 30 min, p65 was observed to translocate into the nucleus from the cytoplasm. However, the treatment with 2 μM nitrobenzoyl sesquiterpenoid 4 substantially abrogated the nuclear translocation of p65, as most of the p65 still stayed in the cytoplasm. These results clearly demonstrated that nitrobenzoyl sesquiterpenoid 4 significantly inhibited NF‐κB activation in the classical pathway by blocking the IκBa phosphorylation and NF‐κB p65 translocation during RANKL‐induced osteoclastogenesis in vitro.

In the alternative pathway of NF‐κB activation, RelB translocates into nucleus and persists the activation of NF‐κB (Abu‐Amer, 2013). We also observed the effect of nitrobenzoyl sesquiterpenoid 4 on the nuclear RelB protein level by western blot. The results revealed that nitrobenzoyl sesquiterpenoid 4 had suppression on the expression level of nuclear RelB protein.

In summary, nitrobenzoyl sesquiterpenoid 4 could interfere with the NF‐κB signalling pathway to reduce osteoclastogenesis through multiple routes.

3.5. Nitrobenzoyl sesquiterpenoid 4 has little influence on RANKL‐induced nuclear expression of c‐Fos

NF‐κB activation consequently induces the activation of c‐Fos (subunit of AP‐1), which translocates into the nucleus to interact with the NFATc1 promoter and initiate osteoclast‐related genes expression (Yamashita et al., 2007). Hence, we investigated the effect of nitrobenzoyl sesquiterpenoid 4 on nuclear c‐Fos by western blot in RANKL‐stimulated pre‐osteoclastic RAW264.7 cells. Nuclear c‐Fos protein levels increased obviously after RANKL stimulation during osteoclastogenesis (Figure 5a). However, nitrobenzoyl sesquiterpenoid 4 (0.5, 1 and 2 μM) exhibited no obvious influence on the nuclear expression of c‐Fos protein (Figure 5a,b).

FIGURE 5.

Nitrobenzoyl sesquiterpenoid 4 (NS4) prevents RANKL‐induced NFATc1 but not c‐Fos activation. RAW264.7 cells were treated with NS4 (0.5, 1 or 2 μM) for 4 h, followed by the stimulation with 100 ng·ml⁻¹ of RANKL for 24 h. Nuclear proteins were then extracted and analysed by western blotting using antibodies against c‐Fos, lamin A/C, β‐actin and NFATc1. The relative protein expression of c‐Fos (a, b) and NFATc1 (e, f) to lamin A/C were determined using ImageJ software. RAW264.7 cells, which had been transfected with NFATc1 luciferase reporter construct, were treated with NS4 (0.5–4 μM) and cyclosporin A (CsA; 1 μM) for 4 h, and then stimulated by 100 ng·ml⁻¹ of RANKL. After 24 h, the luciferase activity (c) was detected. Expression of NFATc1 mRNA (d) followed by RANKL stimulation for 24 h was examined. Values are expressed as the means ± SD (n = 5 independent experiments). ^# P < 0.05, relative to untreated controls, *P < 0.05 relative to RANKL‐treated controls

3.6. Nitrobenzoyl sesquiterpenoid 4 prevents NFATc1 activation induced by RANKL

NF‐κB can bind to the NFATc1 promoter and up‐regulate the expression of NFATc1 (Yamashita et al., 2007), which is believed to be the most distal and critical transcription regulator during RANKL‐induced osteoclast differentiation (Takatsuna et al., 2005; Zhou et al., 2016). Here, we used NFATc1 luciferase reporter assay, real‐time PCR and western blot analysis to examine the effects of nitrobenzoyl sesquiterpenoid 4 on NFATc1 activation in pre‐osteoclastic RAW264.7 cells. First, NFATc1 luciferase activity activated by RANKL was significantly restrained by nitrobenzoyl sesquiterpenoid 4 in a dose‐dependent manner at 0.5 to 4 μM concentrations (Figure 5c). The mRNA expression of NFATc1 induced by RANKL was also significantly inhibited after the treatment of nitrobenzoyl sesquiterpenoid 4 using real‐time PCR (Figure 5d). Furthermore, western blot analysis revealed that treatment with nitrobenzoyl sesquiterpenoid 4 had also remarkably suppressed the expression level of nuclear NFATc1 protein (Figure 5e,f). These results made clear that nitrobenzoyl sesquiterpenoid 4 attenuated osteoclast differentiation by reducing NFATc1 activation.

3.7. Nitrobenzoyl sesquiterpenoid 4 inhibits the expression of RANKL‐induced DC‐STAMP and osteoclast‐related genes

Various osteoclast‐related genes involved in the differentiation and bone resorption of osteoclasts are expressed after NFATc1 activation (Takayanagi et al., 2002). Therefore, we examined the influence of nitrobenzoyl sesquiterpenoid 4 on the expression of osteoclast‐related genes after RANKL induction during osteoclastogenesis. The results showed that nitrobenzoyl sesquiterpenoid 4 dose‐dependently inhibited the mRNA expression of osteoclast‐related genes including TRAP and β3‐integrin in pre‐osteoclastic RAW264.7 cells (Figure 6a). In addition, the mRNA level of a key fusion protein DC‐STAMP in RANKL‐induced osteoclastogenesis was dramatically reduced after 0.5 to 2 μM nitrobenzoyl sesquiterpenoid 4 administration (Figure 6b). DC‐STAMP protein mediates cell–cell fusion involved in osteoclast formation and functions. As a next step, the effect of nitrobenzoyl sesquiterpenoid 4 treatment on DC‐STAMP protein expression by western blot was evaluated. Nitrobenzoyl sesquiterpenoid 4 also showed inhibition on the DC‐STAMP protein stimulated by RANKL in RAW264.7 cells (Figure 6c, d). These results further confirmed the suppressive activity of nitrobenzoyl sesquiterpenoid 4 on the differentiation and bone resorption of osteoclasts.

FIGURE 6.

Nitrobenzoyl sesquiterpenoid 4 (NS4) suppresses RANKL‐induced osteoclast‐related genes and DC‐STAMP. RAW264.7 cells were treated with NS4 (0.5, 1, and 2 μM) for 4 h, and then stimulated by 100 ng·ml⁻¹ of RANKL for 24 h. The expression of osteoclast‐related genes TRAP, β3‐integrin (a) and DC‐STAMP (b) were analysed by using real time‐PCR. Total proteins were extracted and analysed by western blotting after RANKL stimulation for 48 h using antibodies against DC‐STAMP and β‐actin (c, d). The protein expression level of DC‐STAMP relative to β‐actin was established using ImageJ software. Values are expressed as the means ± SD (n = 5 independent experiments). ^# P < 0.05 relative to untreated controls, *P < 0.05 relative to RANKL‐treated controls

3.8. NF‐κB p65 B246 mutation reduces the nuclear translocation of p65 and the inhibition of nitrobenzoyl sesquiterpenoid 4 on NFATc1‐Luc activity during osteoclastogenesis

In the molecular docking assay, we disclosed that the site where nitrobenzoyl sesquiterpenoid 4 bound strongest within NF‐κB p65 protein was residue Arg B246. In order to further value if the effects of nitrobenzoyl sesquiterpenoid 4 on NF‐κB p65 during osteoclastogenesis was due to bind with NF‐κB p65 B246 Arg, we first examined if p65 B246 Arg mutation could functionally affect the translocation of p65 after RANKL stimulation in RAW264.7 cells. NF‐κB p65 protein was over‐expressed in pre‐osteoclastic RAW264.7 cells after constructed WT or mutant NF‐κB p65 plasmid transfection compared with that of the empty vector transfection, showing the successful transfection of both WT and site‐mutant NF‐κB p65 plasmids (Figure 7a,b). After the mutant plasmid NF‐κB p65 R246A transfection, western blot results showed that the NF‐κB p65 protein level in cytoplasm was increased, while NF‐κB p65 nuclear protein level was decreased after RANKL stimulation in RAW264.7 cells compared with that of WT plasmids transfection (Figure 7c–f). Next, we also found that NF‐κB p65 R246A plasmid transfection reduced the RANKL‐induced NFATc1 luciferase activity (Figure 7g). These results demonstrated that site mutation of p65 B246 Arg down‐regulated the nuclear translocation of NF‐κB p65 and NFATc1 transcription during osteoclastogenesis. In addition, the inhibition rate of nitrobenzoyl sesquiterpenoid 4 on RANKL‐induced NFATc1 luciferase activity was significantly reduced after NF‐κB p65 R246A plasmid transfection (21%) compared with that of wt control plasmid (41%) (Figure 7g). Altogether, the above results indicated that the mutation of NF‐κB p65 B246 from Arg to Ala affected the biological functions of p65 and attenuated the inhibitory effects of nitrobenzoyl sesquiterpenoid 4 on NFATc1 signalling pathway, which finally contributed to osteoclastogenesis.

FIGURE 7.

NF‐κB p65 R246A mutation reduces the nuclear translocation of p65 and the inhibition of nitrobenzoyl sesquiterpenoid 4 (NS4) on NFATc1‐Luc activity during osteoclastogenesis. RAW264.7 cells, transfected with NF‐κB p65 R246A (mut) or control WT plasmids (wt) or empty vector, were stimulated with or without RANKL (100 ng·ml⁻¹) for 30 min. Proteins were extracted and analysed by western blotting using antibodies against β‐actin, p65 and lamin A/C. The expression levels of total p65 protein (a, b), cytoplasmic p65 protein to β‐actin (c, d), or nuclear p65 protein level (e, f) to lamin A/C were determined using ImageJ software. RAW264.7 cells, which stably transfected with a NFATc1 luciferase reporter construct, were transiently transfected with NF‐κB p65 R246A or control WT plasmids and stimulated by 100 ng·ml⁻¹ of RANKL. After 24 h, the NFATc1 luciferase activity (g) was detected. Values are expressed as the means ± SD (n = 5 independent experiments). ^# P < 0.05,relative to unstimulated controls, *P < 0.05, relative to vector or RANKL‐treated controls

3.9. Nitrobenzoyl sesquiterpenoid 4 inhibits inflammatory bone loss in vivo

LPS can induce inflammatory bone loss in vivo. This model is usually utilized to value the effects of osteoclast differentiation inhibitors in vivo (Zhai et al., 2014). To evaluate the effect of nitrobenzoyl sesquiterpenoid 4 on inflammatory bone loss in vivo, ICR mice intraperitoneally injected with LPS were treated with nitrobenzoyl sesquiterpenoid 4 intragastrically with no obvious adverse or fatal events being observed during the administration of nitrobenzoyl sesquiterpenoid 4. The left femurs of the animals were removed for micro‐CT analysis. LPS‐injected (5 mg·kg⁻¹) model mice showed extensive decreases in trabecular bone density and quantity (Figure 8a). The quantitative analysis of bone parameters demonstrated significant reductions in BMD, BV/TV and ConnD in the LPS group (Figure 8b). Interestingly, treatment with nitrobenzoyl sesquiterpenoid 4 (1 and 5 mg·kg⁻¹ per day) for 8 days recovered LPS‐induced bone loss by significantly increasing BMD, BV/TV, Tb. N and ConnD, while decreasing Tb.Sp. H&E and TRAP staining, further demonstrating the protective effects of nitrobenzoyl sesquiterpenoid 4 on bone destruction induced by LPS (Figure 8c). Consistent with the micro‐CT results, BV/TV was increased while OC number and OC per bone surface (Oc.S/BS) were significantly reduced after nitrobenzoyl sesquiterpenoid 4 treatment in LPS‐induced inflammatory mice in vivo (Figure 8d–f). In addition, we observed that LPS injection could reduce bone formation in vivo and 5 mg·kg⁻¹ per day nitrobenzoyl sesquiterpenoid 4 treatment could prevent the decrease of bone formation in calcein labelling experiment (Figure 8g,h). This was also in line with the increased bone mass after nitrobenzoyl sesquiterpenoid 4 treatment. Above all, oral administration of nitrobenzoyl sesquiterpenoid 4 protected against inflammatory bone loss and bone destruction in vivo.

FIGURE 8.

Nitrobenzoyl sesquiterpenoid 4 (NS4) prevents inflammatory bone loss induced by LPS in vivo. Groups of mice injected with PBS (n = 6) or LPS (n = 6) but without treatment with NS4 or injected with LPS and treated with 1 mg·kg⁻¹ (n = 6) or 5 mg·kg⁻¹ NS4 (n = 6) were used. Representative 3D reconstructions of transverse (a, above) and longitudinal (a, below) sections of femur from each group by micro‐CT were shown. The parameters of trabecular bone including BMD, BV/TV, Tb. Sp, Tb. N, Cor. Th, Tb. Th and ConnD (b) were analysed. Sections of femur from each group were stained with H&E (c, above) and TRAP (c, middle and below). BV/TV (d), number of osteoclasts (e), and osteoclast surface/bone surface (f) were analysed. In addition, three groups (n = 5) of mice were treated with PBS, LPS and NS4 (5 mg·kg⁻¹) as described above and i.p. injected calcein 7 and 3 days prior to killing. Bone formation indicated by calcein staining was examined by fluorescence microscopy (g) and MAR was calculated (h). ^# P < 0.05 relative to PBS treated controls, *P < 0.05 relative to LPS treated controls

4. DISCUSSION

Osteolytic diseases, such as osteoporosis, metastatic cancers and inflammatory arthritis are characterized by excessive activation of osteoclasts and affects hundreds millions of people worldwide (Boyce, 2013; De Vlam et al., 2014). The osteoclast has been emerging as a new target for treating these osteolytic diseases (Menshawy et al., 2018; Ono & Nakashima, 2018). Currently, bisphosphonate zoledronic acids and denosumab are commonly applied clinically to target osteoclasts, however both of therapies require inconvenient intravenous administration and include the risk of jaw necrosis, which can cause disability (Menshawy et al., 2018; Otto et al., 2018; Wang et al., 2020). Identifying new potential oral inhibitors of osteoclast differentiation is fascinating. Marine sesquiterpenoids display good bioactivities and serve as promising sources of new drugs for development, especially against tumours and inflammation (Cheung et al., 2016; El‐Kassem et al., 2018). In the present study, we evaluated the effects of marine nitrobenzoyl sesquiterpenoids containing a new and special chemical structure on osteoclastogenesis. We found that a nitrobenzoyl sesquiterpenoid, nitrobenzoyl sesquiterpenoid 4, from six nitrobenzoyl sesquiterpenoids and one nitrobenzoyl sesquiterpenoid derivative was the most active compound for suppressing osteoclasts formation and bone absorptive function. Nitrobenzoyl sesquiterpenoid 4 was further observed to inhibit up‐regulated NF‐κB, NFATc1 and DC‐STAMP signals after RANKL activation during osteoclastogenesis in vitro and protect osteolysis in vivo.

The NF‐κB signalling pathway is widely known to play a critical role in osteoclast differentiation (Boyce et al., 1999; Jimi & Ghosh, 2005), NF‐κB p50 and p52 double deficient mice displayed serious osteopetrosis due to osteoclast formation failure (Iotsova et al., 1997). Inhibiting NF‐κB activation in pre‐osteoclastic cells has been employed as a promising therapeutic target for the treatment of osteolytic bone diseases (Zhai et al., 2014; Zhou et al., 2016). Here, NF‐κB‐Luc stably transfected pre‐osteoclastic RAW264.7 cells were first selected and used to screen nitrobenzoyl sesquiterpenoids nitrobenzoyl sesquiterpenoid 1–6 and 7 after RANKL stimulation. Two active compounds, nitrobenzoyl sesquiterpenoid 4 and nitrobenzoyl sesquiterpenoid 1, have nitrobenzoyl moiety connected to C‐14 which is suggested to be the key pharmacophore in the inhibition of NF‐κB‐Luc activities. The more active compound, nitrobenzoyl sesquiterpenoid 4, harbours a hydroxyl group at C‐9 which may further promote the inhibition of NF‐κB activation. This structure–activity relationship was in line with the molecular docking analysis of the nitrobenzoyl sesquiterpenoids with the NF‐κB p65 protein. Additionally, in osteoclastic TRAP assay in vitro, the results verified that nitrobenzoyl sesquiterpenoid 4 was the most active compound identified for suppressing osteoclastogenesis. Then, we focused on nitrobenzoyl sesquiterpenoid 4 and found that it could inhibit both osteoclast differentiation and the resorptive function beginning at 500 nM in vitro.

The above results indicated that nitrobenzoyl sesquiterpenoid 4 could suppress osteoclastogenesis by inhibiting NF‐κB luciferase activity and binding to the NF‐κB protein, which could then serve as a target for nitrobenzoyl sesquiterpenoid 4 to inhibit osteoclastogenesis. So, the mechanisms of nitrobenzoyl sesquiterpenoid 4 on NF‐κB pathway were further revealed. There are classical and alternative pathways in NF‐κB activation. In the classical pathway, IκBa retains p65 and p50 heterodimer in the cytoplasm (Abu‐Amer, 2013; Hayden & Ghosh, 2004; Jimi & Ghosh, 2005). After stimulation by RANKL, IκBa phosphorylates and degrades to release p65 to translocate into nucleus. Here, we observed that nitrobenzoyl sesquiterpenoid 4 suppressed IκBa phosphorylation and nuclear protein levels of NF‐κB p65 by western blot and restrained NF‐κB p65 translocation from the cytoplasm to the nucleus by confocal immunofluorescence assay. Changes in protein structure are essential for the phosphorylation of IκBa and the phosphorylation and translocation of NF‐κB. Inhibitors that bind to the NF‐κB protein can effectively arrest the phosphorylation and the nuclear translocation of NF‐κB (Gilmore & Herscovitch, 2006; Li et al., 2018; Perkins, 2007). In the molecular docking assay, the region where nitrobenzoyl sesquiterpenoid 4 that binds to NF‐κB p65 protein (hydrogen bond with residue Arg B246, Lys B221) is close to the phosphorylation site of NF‐κB p65 protein (Ser‐276) and in the homology region of NF‐κB p65 protein, which involves in the dimerization of NF‐κB members, nuclear translocation and DNA binding (Jacobs & Harrison, 1998; Napetschnig & Wu, 2013; Zhong, SuYang, Erdjument‐Bromage, Tempst, & Ghosh, 1997; Zhong, Voll, & Ghosh, 1998). Therefore, the binding of nitrobenzoyl sesquiterpenoid 4 to NF‐κB p65 protein may affect the phosphorylation of p65, decrease its dimerization with p50, c‐Rel or other NF‐κB members, and also inhibit its binding with the target gene promoters. In this paper, we further found that site mutation of NF‐κB p65 B246 from Arg to Ala reduced the nuclear translocation of p65 and NFATc1 transcription during osteoclastogenesis. NF‐κB p65 B246 can directly serve as a target for nitrobenzoyl sesquiterpenoid 4 to exhibit inhibitory effects on NF‐κB p65 signalling pathway.

The rapid response of classical pathway is followed by a slower response of the alternative pathway, NF‐κB RelB which heterodimerized with either p52 or p50 translocates into nuclear and causes a persistent activation of NF‐κB for several days (Abu‐Amer, 2013; Jimi & Ghosh, 2005). In addition, we also found that nitrobenzoyl sesquiterpenoid 4 decreased the nuclear RelB protein level. Taken together these data suggested that nitrobenzoyl sesquiterpenoid 4 could interfere with the NF‐κB signalling pathways, it could reduce not only IκBa phosphorylation, NF‐κB p65 nuclear translocation in the classical pathway, but also NF‐κB RelB nuclear translocation in the alternative pathway to impair the differentiation and bone resorptive activity of osteoclasts.

Activation of NF‐κB can lead to the up‐regulation of c‐Fos expression during RANKL‐induced osteoclastogenesis (Yamashita et al., 2007). c‐Fos participates in the formation and resorptive function of osteoclasts and c‐Fos knockout mice exhibited severe osteopetrosis with reduced osteoclasts (Matsuo et al., 2000). However, nitrobenzoyl sesquiterpenoid 4 has no action on nuclear c‐Fos protein levels from 0.5 to 2 μM, even though it markedly inhibits the activation of NF‐κB. Our findings suggest that c‐Fos might not be the downstream signal of NF‐κB activation targeted by nitrobenzoyl sesquiterpenoid 4 to treat RANKL‐induced osteoclast differentiation.

NF‐κB activation consequently induces NFATc1, a master transcriptional factor in the terminal differentiation of osteoclasts (Zhai et al., 2014; Zhou et al., 2016). NF‐κB and other nuclear factors can be recruited to the NFATc1 promoter, leading to the autoamplification and activation of NFATc1 (Zhao, Wang, Liu, He, & Jia, 2010). We revealed that nitrobenzoyl sesquiterpenoid 4 not only markedly inhibited RANKL‐induced NFATc1 luciferase activity but also decreased the mRNA and nuclear protein levels of NFATc1 from 0.5 to 2 μM. Interestingly, the inhibition of nitrobenzoyl sesquiterpenoid 4 on RANKL‐induced NFATc1 luciferase activity was attenuated after the mutation of Arg to Ala at drug binding site B246 of NF‐κB p65 protein. Together, the reduced NFATc1 activation by nitrobenzoyl sesquiterpenoid 4 could be partly due to the decreased NF‐κB activation at earlier stage in RANKL‐induced osteoclastogenesis.

Once NFATc1 is activated, it initiates osteoclast‐related genes such as TRAP and β3‐integrin, which subsequently induces the formation and maturation of osteoclasts (Boyle et al., 2003). In the present study, the expression of RANKL‐stimulated osteoclast‐related genes including TRAP and β3‐integrin expressions were all down‐regulated after nitrobenzoyl sesquiterpenoid 4 treatment. Meanwhile, our findings displayed that nitrobenzoyl sesquiterpenoid 4 suppressed both mRNA and protein expression of DC‐STAMP in RANKL‐induced osteoclastogenesis. DC‐STAMP, an essential molecule for pre‐osteoclasts fusion and giant cell formation, determines osteoclast size and enhances its resorptive capacity. DC‐STAMP deficient mice exhibited decreased osteoclast formation and reduced bone‐resorbing activity. DC‐STAMP has also been reported to increase after NF‐κB activation (Chiu & Ritchlin, 2016; Yagi, Miyamoto, Toyama, & Suda, 2006). Following decreased NF‐κB and NFATc1 translocation after nitrobenzoyl sesquiterpenoid 4 treatment during osteoclastogenesis, the suppression of DC‐STAMP could successfully disrupt cell fusion and the formation of functional mature osteoclasts to restrain bone absorption.

LPS is an effective endotoxin that has been regarded as a critical factor for the development of osteolytic bone loss (Zhai et al., 2014). We employed LPS‐induced osteolysis in ICR mice to assess the effects of nitrobenzoyl sesquiterpenoid 4 on osteoclast function in vivo. Interestingly, we found that 1 and 5 mg·kg⁻¹ nitrobenzoyl sesquiterpenoid 4 protected against LPS‐induced bone loss by significantly decreasing osteoclasts and increasing the bone parameter BMD, BV/TV, Tb. N, and ConnD. Osteoclast is a unique cell to absorb bone. These results indicated that oral nitrobenzoyl sesquiterpenoid 4 also attenuated the bone absorptive function of osteoclasts in vivo.

Marine‐derived sesquiterpenoids provide an available resource for drug development (Cheung et al., 2016; El‐Kassem et al., 2018). Here, we demonstrated the new pharmacological function of marine‐derived sesquiterpenoids in osteoclasts, which should extend their future application to treat osteoclast‐related bone lytic diseases. We also identified NF‐κB as a powerful target for potential marine‐derived nitrobenzoyl sesquiterpenoids to suppress osteoclast differentiation. While the NF‐κB signalling pathway was this study's primary focus for nitrobenzoyl sesquiterpenoid 4 to inhibit osteoclast differentiation, it does not exclude that nitrobenzoyl sesquiterpenoid 4 and other nitrobenzoyl sesquiterpenoids can suppress osteoclastogenesis via other targets. We hope to utilize more pre‐osteoclastic cell lineages including human macrophages to evaluate the effects and mechanisms of nitrobenzoyl sesquiterpenoid 4. Moreover, further studies aimed at effects of nitrobenzoyl sesquiterpenoid 4 on osteoblasts which may allow a better understanding of this chemical in its treatment of osteolytic diseases. In addition, we have reported that nitrobenzoyl sesquiterpenoids had anti‐tumour cell proliferation and anti‐inflammation effects in vitro (Tan et al., 2018). Thus, nitrobenzoyl sesquiterpenoid 4 might inhibit both osteoclasts and inflammation to prevent LPS‐induced bone loss in vivo, ovariotomy or tumour‐induced bone loss which is mediated by osteoclasts can also be used to access the effects and mechanisms of action of nitrobenzoyl sesquiterpenoid 4 in vivo. Taken together, multiple topics deserve to be explored in depth in the future.

Collectively, these data showed that the marine fungus‐derived natural nitrobenzoyl sesquiterpenoid, nitrobenzoyl sesquiterpenoid 4 suppressed the RANKL‐induced osteoclasts formation and bone resorption by targeting NF‐κB, along with some data on the structure‐activity relationship of this target. The subsequent mechanistic study showed that nitrobenzoyl sesquiterpenoid 4 could block IκBa phosphorylation, NF‐κB p65 and NF‐κB RelB nuclear translocation, and then decrease NFATc1 activation and DC‐STAMP expression rather than c‐Fos in RANKL‐induced osteoclastogenesis in vitro. Meanwhile, NF‐κB p65 B264 Arg was demonstrated as a novel target for chemical nitrobenzoyl sesquiterpenoid 4 to suppress NF‐κB p65 activation during RANKL‐induced osteoclastogenesis. Additionally, oral nitrobenzoyl sesquiterpenoid 4 attenuated inflammatory bone loss in vivo. Overall this study demonstrated that nitrobenzoyl sesquiterpenoid 4 is a promising lead compound as an osteoclast differentiation inhibitor for the treatment of osteolytic diseases.

AUTHOR CONTRIBUTIONS

X.L. and Y.T. did the study conception and design. Y.T., W.D., Y.Z., B.Z., M.K., X.L., J.S., Y.W., and J.X. performed the experiments and analysed the data. Y.T., K.S.N., Y.L., X.Z., and X.L. drafted and revised the work for intellectual content and context. X.Z. and X.L are responsible for the published work.

CONFLICT OF INTEREST

The authors declare no conflicts interest.

DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for Design and Analysis, Immunoblotting and Immunochemistry and Animal Experimentation, and as recommended by funding agencies, publishers and other organisations engaged with supporting research.

Supporting information

Figure S1. Cell viability of NSs on RAW264.7 cells. RAW264.7 cells (1×10³ cells mL‐1) were cultured with NS1‐3, NS5‐6 or 7 (0, 0.5, 1, 2 μM) for 5 d, and the cell viability of NS1‐3, NS5‐6 or 7 on RAW264.7 were evaluated by MTT assay kit. Values are expressed as the means ± SD of independent experiments (n = 5). # p < 0.05 relative to untreated controls.

Figure S2. Molecular docking of NS1‐3, NS5‐6 and 7 with NF‐κB p65. Binding sites of the molecule (A) NS1, (C) NS2, (E) NS3, (G) NS5, (I) NS6 and (K) 7 with NF‐κB p65 protein. Hydrophobic, polar and the exposed regions of the receptor are depicted in green, purple, and red colors, respectively. The interaction details of the predicted binding mode of (B) NS1, (D) NS2, (F) NS3, (H) NS5, (J) NS6 and (L) 7 with p65, respectively. The contact residues are displayed and labeled by type and number, along with detailed interaction types, distance and energy.

Figure S3. Cell toxicity of NS4 on RAW264.7 cells and BMMs. RAW264.7 cells or BMMs were treated with NS4 (0, 1, 2 μM) for 96 h, 72 h respectively. After washed with precooled PBS, cells were suspended in 500 μL binding buffer and stained with PI for 30 min in the dark at room temperature. Then the PI positive dead cells were captured and recorded by FACS Canto II (BD, Triangle, NC, USA). RAW264.7 cells or BMMs were treated in 70 °C water‐bath for 5 min as a positive control respectively. Results were analyzed with FLOWJO VX software and expressed as the percentage of PI positive cells of (A) RAW264.7 cells and (B) BMMs. In addition, (C) RAW264.7 cells (1×10³ cells mL‐1) or (D) BMMs (1×10⁵ cells mL‐1) were treated with NS4 (0, 1, 2 μM) for 96 or 72 h respectively, then the supernatants were collected to detect LDH (a cytosolic enzyme released upon membrane damage in necrotic cells) followed by the manufacturer’s instructions. Values are expressed as the means ± SD of independent experiments (n = 5). ### p < 0.001 relative to untreated controls.

Figure S4. Effects of NS4 on the proliferation of MC3T3‐E1 osteoblast cells. MC3T3‐E1 cells (1×10³ cells mL‐1) were cultured with NS4 (0, 0.5, 1, 2 μM) for 5 d, and the effects of NS4 on MC3T3‐E1 cell proliferation were evaluated by MTT assay kit. Values are expressed as the means ± SD of independent experiments (n = 5). # p < 0.05 relative to untreated controls.

Figure S5. Molecular docking of NS4 with IκBa protein and the effects of IκBa Lys 132 mutation on the phosphorylation of IκBa. (A) Binding sites of the molecule NS4 with IκBa protein. Hydrophobic, polar and the exposed regions of the receptor are depicted in green, purple, and red colors. (B) The interaction details of the predicted binding mode of NS4 with IκBa. The contact residues are displayed and labeled by type and number, along with detailed interaction types, distance and energy. (C) The transfection efficiency of IκBa K132A plasmids was detected by fluorescence microscopy which indicated by the co‐expressed GFP. (D‐E) RAW264.7 cells, transfected with IκBa K132A (mut) or control plasmids (wt) or empty vector, were stimulated with or without RANKL (100 ng mL‐1) for 30 min. Proteins were extracted and analyzed by western blotting using antibodies against IκBa, p‐IκBa, β‐actin. The relative protein expression levels of IκBa or p‐IκBa to β‐actin, were determined using ImageJ software. Values are expressed as the means ± SD (n = 5 independent experiments). *** p < 0.001 relative to vector. ## p < 0.01 relative to untreated control.

Tan Y, Deng W, Zhang Y, et al. A marine fungus‐derived nitrobenzoyl sesquiterpenoid suppresses receptor activator of NF‐κB ligand‐induced osteoclastogenesis and inflammatory bone destruction. Br J Pharmacol. 2020;177:4242–4260. 10.1111/bph.15179