A total of 12 10-week-old male Sprague Dawley rats (Laboratory Animal Center of Ning Xia Medical University, Yinchuan, China; protocol no. 2020-0001) weighing 294±13 g, were group-housed at a constant temperature (25°C), at 55% humidity in a 12-h light/dark cycle with food and water available ad libitum. Bone marrow-derived macrophages (BMMs) and BMSCs were isolated from whole bone marrow of the rats as described previously (27). Briefly, bone marrow cells were flushed out of the medullary cavity of the femur and tibia using α modified Eagle's medium (α-MEM, cat. no. 01-042-1ACS; Biological Industries) containing 2% FBS (cat. no. 10099141) and 1% penicillin/streptomycin (cat. no. 15140122; both from Thermo Fisher Scientific, Inc.), erythrocytes were removed with lysis buffer (cat. no. R1010; Beijing Solarbio Science & Technology Co., Ltd.), and the remaining cells were resuspended in α-MEM containing 10% FBS, 1% penicillin/streptomycin and 30 ng/ml macrophage colony-stimulating factor (M-CSF, cat. no. 216-MC-025; R&D Systems, Inc.) and incubated at a density of 1×10⁶ cells/ml in a 10-cm dish for 24 h. Non-adherent cells were considered as BMMs, which were centrifuged (1,000 × g, 5 min, 20°C) and reseeded in another 10-cm dish and incubated in an atmosphere of 5% CO₂ at 37°C for a further 3-4 days until the cells reached 90% confluence. At the same time, adherent cells continued to be cultured in α-MEM containing 10% FBS and 1% penicillin/streptomycin, and the culture medium was replaced every 2 days until a large number of BMSCs were obtained by cell passage. RAW264.7 cells (American Type Culture Collection) were cultured in α-MEM containing 10% FBS and 1% penicillin/streptomycin in a humidified atmosphere of 5% CO₂ at 37°C and the medium was replaced every 2 days.

BMMs were plated into a 96-well plate at a density of 6×10³ cells/well and incubated overnight, and subsequently the medium was replaced with fresh α-MEM containing 30 ng/ml M-CSF, 50 ng/ml RANKL (cat. no. 390-TN-010; R&D Systems, Inc.), and various concentrations (0, 0.5, 1, 2.5 or 5 µM) of DHA (cat. no. D7439; MilliporeSigma). The α-MEM was replaced every 2 days until mature osteoclasts had formed. Cells were stained for TRAP using a kit (cat. no. PMC-AK04F-COS; Cosmo Bio Co., Ltd.), following the manufacturer's instructions. Briefly, cells were washed with PBS three times (5 min each wash), fixed with 4% paraformaldehyde at 37°C for 15 min, and incubated in the presence of the TRAP solution for 5 min at 20°C. TRAP-positive cells with more than three nuclei were identified as mature osteoclasts and counted, and the number and area of osteoclasts in five fields of view per well were measured using the Olympus DP71 light microscope (Olympus Corporation) (24).

Osteogenic differentiation was detected when BMSCs were cultured to 3-5 generations. In brief, BMSCs were plated at a density of 1×10⁵ cells/well into collagenase-treated 6-well plates (cat. no. 354400; Corning, Inc.) and incubated overnight at 37°C, and then simulated with osteogenic medium (α-MEM containing 10% FBS, 1% penicillin/streptomycin, 10 nM dexamethasone, 10 mM β-glycerophosphate and 50 µg/ml ascorbate; cat. no. RASMX-90021; Cyagen Biosciences, Inc.) and various concentrations (0, 0.5, 1, 2.5 and 5 µM) of DHA for 7 and 21 days, with the medium being replaced with fresh medium every 2 days. The alkaline phosphatase (ALP) activity and mineralization nodules of BMSCs were observed following fixation with 4% paraformaldehyde at 20°C for 15 min and staining with an ALP detection kit (cat. no. SCR004; MilliporeSigma) in the dark at 20°C for 10 min and 1% alizarin red (cat. no. G1038-100ML; Wuhan Servicebio Technology Co., Ltd.) at 20°C for 10 min. ImageJ 1.48v software (National Institutes of Health) was used to measure the percentages of the mineralized area and ALP-positive cells (28).

CCK-8 kit (cat. no. C0038; Dojindo Laboratories, Inc.) was used to determine the effects of DHA on the proliferation and viability of BMMs and BMSCs, according to the manufacturer's instructions. Cells were seeded in 96-well plates at a density of 6×10³ cells/well and incubated overnight. α-MEM containing various concentrations (0, 0.5, 1, 2.5, 5 and 10 µM) of DHA were added, and cells were incubated for 96 h at 37°C. Subsequently, 10 µl CCK-8 buffer was added to each well, and the cells were incubated for a further 2 h at 37°C. The optical density (OD) was then measured at 450 nm on a Multiskan absorbance microplate reader (Thermo Fisher Scientific, Inc.). Cell viability was calculated according to the following formula: Cell viability=(experimental group OD-zeroing OD)/(control group-OD zeroing OD) (24).

Tetraethyl rhodamine isothiocyanate (TRITC)-conjugated phalloidin [cat. no. 40734ES75; Yeasen Biotechnology (Shanghai) Co., Ltd.] was used to detect the F-actin rings in osteoclasts, following the manufacturer's instructions. Briefly, BMMs were seeded into 24-well plates at a density of 2×10⁴ cells/well. Cells were cultured to adhere to the plate overnight at 37°C, and were subsequently cultured with medium containing different concentrations (0, 1 and 5 µM) of DHA, 30 ng/ml M-CSF and 50 ng/ml RANKL. The medium was replaced every 2 days. After 6 days of culture, cells were washed thrice with preheated (37°C) PBS (5 min each wash), followed by fixation with 4% paraformaldehyde at 20°C for 10 min, and washed three times with PBS again (5 min each wash). Subsequently, the cells were permeabilized with 0.5% (v/v) Triton X-100 (cat. no. 9002-93-1; MilliporeSigma) for 5 min, and washed three times with PBS again (5 min each wash). The cells were then incubated with 200 nM TRITC-conjugated phalloidin diluted in 0.5% (w/v) Invitrogen^® BSA (cat. no. D1205; Thermo Fisher Scientific, Inc.)-PBS for 1 h at 37°C, washed three times with PBS (5 min each wash), and fluorescence quenching resistant sealing tablets containing DAPI (cat. no. ZLI-9557; Beijing Zhongshan Jinqiao Biotechnology Co., Ltd.) were used for staining the nuclei at 20°C for 10 min. The F-actin ring distribution was observed using an LSA1 confocal microscope (Nikon Corporation). Finally, the fluorescence images were processed using NIS-Elements AR 4500 software (Nikon Corporation), and the number of intact F-actin rings was counted using ImageJ 1.48v software (National Institutes of Health).

Bone resorption of the osteoclasts was determined using a resorption assay system using cell cultures on bone slices (cat. no. DT-1BON1000; Immunodiagnostic Systems, Ltd.). BMMs were seeded in 96-well plates with bone slices at a density of 6×10³ cells/well. Cells were first cultured overnight at 37°C to adhere to the plates, and subsequently were cultured with α-MEM containing different concentrations (0, 1 and 5 µM) of DHA, 30 ng/ml M-CSF and 50 ng/ml RANKL, the medium was replaced every 2 days. After 8 days of culture, the bone slices were rinsed in PBS, fixed in 2.5% glutaraldehyde at 4°C for 7 min, placed in 0.25 M ammonium hydroxide and sonicated three times with 50 Hz (5 min each sonication) to remove the cells. Bone slices were washed in distilled water and dehydrated by passing through graded (40, 75, 80, 95 and 100%) alcohol for 10 min. After being air-dried, bone slices were stained with 0.5% (w/v) aqueous toluidine blue (cat. no. G3668; Beijing Solarbio Science & Technology Co., Ltd.) at 20°C for 3 min, and the bone resorption areas were subsequently examined under a light microscope. To detect resorption pits, bone slices were fixed in 2.5% glutaraldehyde at 20°C for 2 h, then fixed in 1% osmic acid at 20°C for a further 2 h, then washed in distilled water and dehydrated by passing through graded (40, 75, 80, 95 and 100%) alcohol for 10 min. The bone slices were then dried by passing through graded (75, 100%) tert-butyl alcohol. Following which, bone slices were sputtered with gold in an airless spray unit, and the presence of resorption pits on bone slices was observed using an S-3400N scanning electron microscope (Hitachi, Ltd.). Finally, the number of resorption pits was measured using ImageJ 1.48v software (National Institutes of Health) (29).

RAW264.7 cells were transfected with the pGL4.32 (luc2P/NF-κB-RE/Hygro) vector (NF-κB luciferase reporter plasmid, 0.5 µg, cat. no. E8491) and the pGL4.30 (luc2P/NFAT-RE/Hygro) vector (NFAT luciferase reporter plasmid, 0.5 µg, cat. no. E8481; both from Promega Corporation) using the Gene Pulser Xcell™ Electroporation system (cat. no. 165-2660; Bio-Rad Laboratories, Inc.) for stable transfection, according to the manufacturer's instructions, and the following electroporation parameters: i) Voltage, 200 V; ii) capacitance, 950 µF; iii) pulse duration, 5 msec; iv) number of pulses, 1; and v) electroporator buffer, Opti-MEM (cat. no. 51985091; Thermo Fisher Scientific, Inc.). The pSV-β-galactosidase control vector (0.5 µg, cat. no. E1081; Promega Corporation) was co-transfected, according to the manufacturer's instructions, for normalization. Cells transfected with NFATc1, NF-κB and β-galactosidase were selected using 400 µg/ml G418 (cat. no. 10131035; Thermo Fisher Scientific, Inc.) to investigate NFATc1 and NF-κB activation as described previously (30,31). Briefly, the transfected RAW264.7 cells were plated in 48-well plates at a density of 1×10⁵ or 5×10⁵ cells/well in triplicate. After 24 h, the cells were pretreated with various concentrations (0, 0.5, 1, 2.5 and 5 µM) of DHA for 1 h, and the cells were subsequently incubated with α-MEM containing 50 ng/ml RANKL for 6 h to activate NF-κB, and 24 h to activate NFATc1. Cells were rinsed in PBS buffer, then lysed with cell culture lysis reagent (120 µl/well, cat. no. E1531; Promega Corporation) for 10 min at 4°C, before the supernatant was collected by centrifugation (10,000 × g for 25 min) at 4°C. The luciferase activities of NFATc1 and NF-κB were detected using the Luciferase Assay Kit (cat. no. E1483; Promega Corporation) and normalized to the β-galactosidase activity in each sample.

BMMs were plated into 6-well plates at a density of 1×10⁶ cells/well and cultured at 37°C to allow the cells to adhere to the plates. BMMs were pretreated with 5 µM DHA for 4 h, and then stimulated with 30 ng/ml M-CSF and 50 ng/ml RANKL for the indicated time periods (0, 10, 30 and 60 min and 0, 1, 3 and 5 days). Subsequently, the cells were treated for protein extraction following the manufacturer's instructions. Briefly, the medium was removed, and the cells were collected using 0.25% trypsin (Beijing Solarbio Science & Technology Co., Ltd.) digestion at 37°C for 3 min. After centrifugation (1,000 × g, 5 min, 20°C) and washing, cells were lysed using a buffer containing protease inhibitors and phosphatase inhibitors from the whole cell lysis assay kit (cat. no. KGP250; Nanjing KeyGen Biotech Co., Ltd.) at 4°C for 20 min. The protein concentrations were quantified using a BCA protein assay kit (cat. no. KGPBCA; Nanjing KeyGen Biotech Co., Ltd.) after centrifugation (12,000 × g, 10 min, 4°C). The extracted proteins were then boiled in the loading buffer (cat. no. GH101-01; TransGen Biotech Co., Ltd.) for 10 min. Total proteins (30 µg) from each group were separated via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (8 or 12% gels), and then transferred onto polyvinylidene difluoride membranes. After blocking with 5% skimmed milk in Tris-buffered saline containing 0.1% Tween-20 (TBST) at 20°C for 2 h, the membranes were incubated overnight at 4°C with primary antibodies against IκBα (1:1,000; cat. no. 51066-1-AP), NF-κB (1:1,000; cat. no. 66535-1-Ig; both from ProteinTech Group, Inc.), phosphorylated (p)-ERK (1:10,000; cat. no. ab229912), ERK (1:10,000; cat. no. ab184699), p-JNK (1:10,000; cat. no. ab4821), JNK (1:10,000; cat. no. ab179461), p-p38 (1:10,000; cat. no. ab4822), p38 (1:10,000; cat. no. ab170099), NFATc1 (1:10,000, cat. no. ab264530; all from Abcam) and β-actin (1:5,000; cat. no. 20536-1-AP; ProteinTech Group, Inc.). After washing three times with TBST (5 min each wash), the membranes were incubated with goat anti-rabbit HRP-conjugated secondary antibodies (1:5,000; cat. no. SA00001-2; ProteinTech Group, Inc.) at 20°C for 1 h. Finally, membranes were soaked in enhanced chemiluminescence reagent (cat. no. KGP1121; Shanghai Beiyi Bioequip Information Co., Ltd.) for 1 min, and imaged. Band intensities were detected, the data were semi-quantified using ImageJ 1.48v software (National Institutes of Health), and are shown relative to the intensity of the β-actin control.

BMMs were plated into 6-well plates at a density of 1×10⁶ cells/well and cultured overnight at 37°C to allow the cells to adhere to the plates. The α-MEM was then replaced with fresh medium containing different concentrations (0, 0.5, 1, 2.5 and 5 µM) of DHA, 30 ng/ml M-CSF and 50 ng/ml RANKL, and cell culture was allowed to continue with the medium being replaced every 2 days. After 6 days of culture, cells were collected and treated for total RNA isolation using Multisource Total RNA Prep kits (cat. no. AP-MN-MS-RNA-250; Axygen; Corning, Inc.) following the manufacturer's instructions. The same procedure was performed in BMSCs cultured for 14 days with osteogenic medium for osteogenic differentiation. Complementary DNA was synthesized from 1 µg total RNA using a TransScript^® All-in-One First-Strand cDNA Synthesis kit (cat. no. AT341-01; TransGen Biotech Co., Ltd.), according to the manufacturer's protocol, and an S1000 Thermal Cycler (Bio-Rad Laboratories, Inc.). qPCR was performed to quantify the expression of the target genes during osteoclast formation and the osteoblastic differentiation of BMSCs using PerfectStart™ Green qPCR SuperMix (cat. no. AQ602-21; TransGen Biotech Co., Ltd.) and an Applied Biosystems 7500 Fast Real-Time PCR System (Thermo Fisher Scientific, Inc.) with the following thermocycling conditions: 45 cycles of 94°C for 30 sec; 54°C for 30 sec; and 72°C for 34 sec. The primer sequences are listed in Table I. qPCR for each group was performed in triplicate, and the expression levels of target genes were quantified using the 2^−ΔΔCq method (32). The mean C_q value in each experimental group was normalized to the control group, and β-actin was used as the quantitative control gene.

A total of 36 10-week-old male Sprague Dawley rats (Laboratory Animal Center of Ning Xia Medical University, Yinchuan; protocol no. 2020-0001) weighing 294±13 g, were group-housed at a constant temperature (25°C) at 55% humidity in a 12-h light/dark cycle with food and water available ad libitum. All 36 rats were randomly divided into three groups, with 12 rats per group: i) The sham-operated group; ii) the destabilization of medial meniscus (DMM) + vehicle group; and iii) the DMM + DHA group. Rats were acclimated for 1 week and anesthetized with 1% pentobarbital sodium in PBS (60 mg/kg). Surgery was performed on the right knee of each rat, as previously described (33). Briefly, the right knee joint capsule of each rat for which DMM surgery was performed was exposed according to a medial parapatellar approach; subsequently, the medial meniscotibial ligament was transected with micro-scissors, and the medial meniscus was reflected proximally towards the femur before suturing the joint capsule and skin with a 5-0 synthetic absorbable suture. By contrast, the rats of the sham-operated group were subjected to the same procedure without transections of medial meniscotibial ligament and medial meniscus. The rats of the DMM + DHA group were subjected to an intraperitoneal injection of DHA at 1 mg/kg every 2 days after the first post-operative day until the time of sacrifice (4 and 8-weeks after the operation). The rats in the sham-operated and DMM + vehicle groups were subjected to an intraperitoneal injection of 1% DMSO as a control. DHA was prepared as described previously in the literature (34). In brief, 25 mg DHA was weighed out and dissolved in 1 ml prepared DMSO, and then diluted in 99 ml sterile PBS solution to prepare a 1 mg/kg DHA. As a vehicle, the sham-operated and DMM + vehicle groups were administered solvent with the same dose, frequency and duration as the DMM + DHA group. A schematic representation of the experimental design is shown in Fig. 1.

All experimental procedures in this study were approved by the Animal Care and Experiment Committee of Ningxia Medical University (approval. no. IACUC-NYLAC-2020-51; Yinchuan, China), and all measures were taken in accordance with the principles and guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals (35) to minimize the number and suffering of rats included in this experiment.

Six rats from each group were fasted, but had free access to water, for 6 h prior to euthanasia. Sodium pentobarbital (1%, 60 mg/kg) was injected intraperitoneally for anesthesia prior to cardiac puncture, and additional sodium pentobarbital (1%, 100 mg/kg) was injected intraperitoneally for euthanasia in the study. Blood collected at 4 and 8 weeks post-operatively was clotted at 20°C for 30 min, and then centrifuged (4,000 × g, 20 min) at 4°C. Serum was then separated and stored at -80°C. The levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were analyzed using ELISA kits (cat. nos. AE90375Ra and AE91382Ra, respectively; Shanghai Lianshuo Biological Technology Co., Ltd.), as described by the manufacturer (28).

In order to observe the effects of DHA on bone remodeling, a Skyscan 1,176 µCT (Bruker Corporation) was used to examine the microstructural changes in the tibial subchondral bone as described previously (36). Briefly, the knee joints of six rats in each group were collected, and excess soft tissue was dissected. Following fixation using 10% neutral formalin buffer for 48 h at 20°C, the cleaned knee joints were imaged by µCT with the following instrument settings: Resolution, 10 µm/pixel; voltage, 40 kV; current, 250 µA; exposure time, 0.9 sec; and rotation, 0.4° step through 180°. A region of interest was identified as the portion (3.5 mm ventrodorsal length, 0.5 mm below the growth plate and 1 mm in height) of the load-bearing region at the medial tibial plateau to use for determination of the bone parameters, including bone volume fraction (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp) and connectivity density (CD).

After fixation using 10% neutral formalin buffer for 48 h at 20°C, the knee joints of six rats in each group were decalcified in 10% disodium ethylenediaminetetraacetate dihydrate (cat. no. E8030; BioInsight) for 4 weeks at 20°C prior to embedding in paraffin, and the joints were sectioned at 5-µm thickness along the coronal plane for staining. H&E staining kit (cat. no. G1005; Wuhan Servicebio Technology Co., Ltd.) was used to calculate the thickness of the hyaline and calcified (CC) cartilage layers, as described previously (37). Briefly, the sections were placed in xylene twice for 15 min, anhydrous ethanol twice (7 min immersion) and finally, in 75% alcohol for 7 min. The sections were stained with hematoxylin for 5 min at 20°C, rinsed in running water for 5 min, dehydrated in 85 and 95% alcohol for 10 min each, and finally stained with eosin for 5 min at 20°C. Before staining with Safranin O-Fast Green and TRAP (G1053 and G1050, respectively; Wuhan Servicebio Technology Co., Ltd.), the slides were deparaffinized as described above, and subsequently stained with Fast Green for 6 min, washed at 20°C, dehydrated, and stained with Safranin O for 3 min at 20°C. The Osteoarthritis Research Society International (OARSI) scores were graded to evaluate the degeneration of articular cartilage, as described previously (38). TRAP staining was performed to measure the number of mature osteoclasts and the ratio of osteoclast surface per bone surface (Oc.S/BS), as described previously (39). Slides were incubated in working fluid derived from the TRAP staining kit for 2 h, washed three times with distilled water, and subsequently stained with hematoxylin for 5 min at 20°C. After covering the slides with neutral resin, five random views from three sections per rat were visualized using a DP71 light microscope with DP controller 3.3.1.292 software (Olympus Corporation) and quantified using ImageJ 1.48v software (National Institutes of Health).

For immunohistochemistry staining, after decalcification and fixation according to the same conditions described above, slides from each treatment group at 4-weeks after the surgery were incubated with 3% hydrogen peroxide for 20 min at 37°C, and then blocked with 5% BSA (Beijing Solarbio Science & Technology Co., Ltd.) for 30 min at 37°C. The slides were subsequently incubated overnight at 4°C with primary antibodies (all diluted 1:100) against RANKL (cat. no. 23408-1-AP), CXCL12 (cat. no. 17402-1-AP; both from ProteinTech Group Inc.) and NFATc1 (cat. no. ab264530; Abcam). After washing three times (5 min each wash), sections were incubated with the goat anti-mouse/rabbit poly-HRP secondary antibody (1:200; cat. no. PR30009; ProteinTech Group, Inc.) for 1 h at 37°C, and subsequently 3,3-diaminobenzidine (DAB) (cat. no. ZLI-9018; Beijing Zhongshan Jinqiao Biotechnology Co., Ltd.) was added to develop the color prior to counterstaining with hematoxylin for 5 min at 20°C. After covering the slides with neutral resin, positively stained cells from three sections per rat were visualized using a DP71 light microscope with DP controller 3.3.1.292 software (Olympus Corporation) and quantified using ImageJ 1.48v software (National Institutes of Health).

Experiments were repeated at least three times, and the data are expressed as the mean ± standard deviation. All statistical analyses were performed using GraphPad Prism 7.0 (GraphPad Software, Inc.). An unpaired Student's t-test was used to compare the parameters between two groups, whereas ANOVA followed by Tukey's post-hoc test was used to compare the parameters among ≥3 groups. P<0.05 was considered to indicate a statistically significant difference.

To ascertain the effects of DHA on the differentiation of BMMs into osteoclasts, the BMMs were first isolated and treated with M-CSF and RANKL in the absence or presence of DHA at various concentrations (0, 0.5, 1, 2.5 or 5 µM). As shown by TRAP staining, BMMs underwent differentiation into mature TRAP-positive multinucleated osteoclasts in the presence of M-CSF and RANKL. On the other hand, the formation of multinucleated osteoclasts was notably suppressed by DHA in a dose-dependent manner (Fig. 2A). The number of osteoclasts (TRAP-positive with >3 nuclei) was 232.00±3.86 per well in the group treated without DHA, and 11.85±2.20 per well in the group treated with 5 µM DHA (Fig. 2B). Additionally, BMM differentiation was observed by TRAP staining at different time points, and osteoclast formation was found to be notably suppressed after 3 days of treatment with 5 µM DHA (Fig. 2C). The numbers and areas of osteoclasts further supported that DHA could exert an inhibitory effect on the early stage of osteoclastogenesis (Fig. 2D).

In order to further explore the underlying mechanism of DHA affecting osteoclast formation, the expression levels of genes associated with osteoclastogenesis and function were then examined. In response to RANKL stimulation, the expression levels of genes, including CTSK, CTR, TRAP, MMP-9, RANK and CXCR4 were high in the group treated without DHA. However, the expression levels of these genes were significantly inhibited in a dose-dependent manner in the groups treated with various concentrations of DHA (Fig. 2E).

To determine whether the inhibitory effect of DHA on osteoclast formation was associated with cytotoxicity, BMM viability assays were performed by using CCK-8. The results obtained demonstrated that DHA did not produce cytotoxicity with BMMs at concentrations below 10 µM (Fig. 3A), at which concentration the formation of osteoclasts was significantly suppressed after 48-96 h of treatment (Fig. 3B-D).

BMSCs have the ability to differentiate into osteoblasts, which are the major source of RANKL, affecting osteoclast formation (40). Therefore, further experiments were designed to investigate whether DHA affected the osteogenic differentiation of BMSCs. Similar to the experiments performed on BMMs, no cytotoxicity was observed with BMSCs at concentrations of DHA up to 5 µM (Fig. 4A). Given that the expression of ALP is a representative phenotypic indicator of the early osteogenic differentiation stage of BMSCs (41), an ALP staining kit was used to evaluate the effect of DHA on ALP activity of BMSCs cultured with osteogenic medium. The results obtained showed that treatment with different concentrations of DHA for 7 days did not change ALP activity (Fig. 4B and D). To evaluate the effect of DHA on calcium mineralization, BMSCs cultured with osteogenic medium for 21 days were stained with alizarin red. As shown in Fig. 4C, DHA did not affect the formation of mineralization nodules in BMSCs cultured with osteogenic medium. Furthermore, RT-qPCR analysis demonstrated that DHA had no effect on the expression of osteoblast marker genes, including Runx family transcription factor 2 (Runx2), Alp and bone γ-carboxyglutamate protein (Bglap) (Fig. 4E). These results suggested that DHA did not affect the osteogenic differentiation of BMSCs.

Given that the formation of F-actin rings is necessary for osteoclastic bone resorption (29), the effect of DHA on F-actin rings was further investigated. F-actin rings with a typical appearance, stained with TRITC-conjugated phalloidin were observed in the group treated without DHA. However, treatment with different concentrations of DHA resulted in marked alterations in the number and morphology of F-actin rings (Fig. 5A). The number of F-actin rings per view was 17.13 in the group treated without DHA, 11.58 in the group treated with 1 µM DHA and 1.33 in the group treated with 5 µM DHA (Fig. 5D). These results suggested that DHA inhibited F-actin ring formation in osteoclasts in vitro.

As bone resorption is associated with the formation of osteoclasts and F-actin rings (29), the effects of DHA on osteoclast bone resorption using bone slices were investigated next. As shown in Fig. 5B and C, large areas were stained with toluidine blue and a number of large bone resorption pits were observed on the bone slices in the group treated without DHA; by contrast, smaller areas stained with toluidine blue and fewer bone resorption pits were observed on the bone slices in the groups treated with DHA. The resorption areas decreased from 51.38 to 33.36% and values <5% after treatment with 1 and 5 µM DHA, respectively (Fig. 5E). Furthermore, the number of resorption pits/mm² was reduced from 94.39 to 10.3 after treatment with 5 µM DHA (Fig. 5F). Taken together, these results suggested that DHA suppressed osteoclast bone resorption in vitro.

In order to further unravel the mechanism involved in the inhibition of DHA on osteoclast formation and bone resorption, RAW264.7 cells stably transfected with NFATc1 and NF-κB luciferase reporter constructs were used to investigate the activities of the two transcription factors. The data revealed that various concentrations (0, 0.5, 1, 2.5 and 5 µM) of DHA significantly inhibited NFATc1 and NF-κB luciferase activities in a dose-dependent manner (Fig. 6A and B).

The two important transcription factors, NFATc1 and NF-κB, and the protein kinases associated with the MAPK signaling pathway (such as ERK, JNK and p38) fulfill important roles in osteoclast formation. Consequently, western blot analysis was performed to examine the protein levels of NFATc1, IκBα, NF-κB, p-ERK1/2, ERK1/2, p-JNK, JNK, p-p38 and p38 in BMMs pretreated with or without 5 µM DHA for 4 h, followed by treatment with 50 ng/ml RANKL for the indicated time periods. As shown by western blot analysis, NFATc1 was significantly reduced upon treatment with 5 µM DHA from 3 to 5 days post-stimulation of RANKL (Fig. 6C and D). Administration of 5 µM DHA significantly inhibited IκBα degradation and NF-κB activation at 30 min post-stimulation of RANKL (Fig. 6E and F). The inhibitory effect of DHA on p-p38 and p-JNK was observed at 30 min post-stimulation of RANKL, whereas its inhibitory effect on p-ERK1/2 was observed at 60 min post-stimulation of RANKL (Fig. 6G and H). Taken together, these results suggested that DHA could suppress RANKL-induced osteoclastogenesis through inhibiting the NFATc1, NF-κB and MAPK signaling pathways.

The DMM-induced OA rats were treated with or without DHA for 4 and 8 weeks and, subsequently, the microstructure of the tibial subchondral bone was examined by µCT (Fig. 7A). The resultant microstructure indices demonstrated that a significant bone loss occurred in the DMM + vehicle group compared with the sham-operated group in terms of decreased BV/TV (Fig. 7B), Tb.N (Fig. 7C) and CD (Fig. 7F), reduced Tb.Th (Fig. 7D) and increased Tb.Sp (Fig. 7E) values at 4 weeks after the surgery. However, the overall trends in these changes in the DMM + DHA group were preserved by an intraperitoneal injection of DHA at 1 mg/kg every 2 days for 4 and 8 weeks, and the levels of the ALT (Fig. 7G) and AST (Fig. 7H) biomarkers indicated that DHA was not toxic to rats.

According to the results of TRAP staining on tibial subchondral bone, it was found that the number of TRAP-positive osteoclasts was higher in the DMM + vehicle group compared with that in the sham-operated group, whereas DHA significantly inhibited osteoclast formation at 4 weeks post-surgery (Fig. 8A and D). These results corroborated the experiments wherein subchondral bone resorption was detected by µCT, supporting the hypothesis that DMM causes subchondral bone loss by enhancing bone resorption mediated by promoting osteoclast formation during the early stage of OA.

H&E and Safranin O-Fast Green staining was subsequently used to assess histomorphological changes in the cartilage during OA progression. It was found that cartilage degeneration generally occurred, and become more severe in a time-dependent manner, after OA induction. H&E staining revealed that the thickness of HC decreased, whereas that of CC increased and moved closer to the articular surface in the DMM + vehicle group compared with the sham-operated group (Fig. 8B). Treatment with DHA, however, significantly attenuated the increasing thickness of CC and the CC/total articular cartilage (TAC) ratio (Fig. 8E).

Safranin O-Fast Green staining showed that, compared with the sham-operated group, the matrix and chondrocytes in the DMM + vehicle group were lost, and irregular cracks had developed at 4 weeks post-surgery; in addition, full-thickness degeneration of articular cartilage occurred and became more widespread by 8 weeks, with the development of OA (Fig. 8C). However, DHA significantly alleviated the cartilage degeneration grade and delayed the onset of OA. The OARSI score results demonstrated that, compared with the sham-operated group, a markedly increased rate of cartilage degeneration occurred in the DMM + vehicle group, whereas cartilage degeneration was suppressed in the DMM + DHA group (Fig. 8F).

RANKL, CXCL12 and NFATC1 are proteins that are known to affect the formation and migration of osteoclasts (16,42), and immunohistochemical staining was performed to assess the effect of DHA on these proteins. The results revealed that significantly more cells in the DMM + vehicle group were stained positively for RANKL, CXCL12 and NFATc1 (Fig. 9A-D) compared with the sham-operated group; however, DHA suppressed the increased expression levels of RANKL, CXCL12 and NFATc1 induced by DMM (Fig. 9A-D). These findings further corroborated the inhibitory effect of DHA on osteoclast formation and function.