Fucoidan (Sigma-Aldrich; Merck Millipore, Darmstadt, Germany) was digested consecutively with glucoamylase, pectinase and cellulose complex (Sigma-Aldrich; Merck Millipore), into three different LMWF solutions as previously described (18). Fucoidan was dissolved in dimethyl sulfoxide (10 mg/ml), filtered using a 0.2 mm filter and stored at −20°C. The effect of these LMWF solutions on the growth and viability of osteoclasts was evaluated using MTT assay as previously described (19), in order to select the strongest inhibitory concentration as preparation for the subsequent experiments. The concentrations of the three LMWF solutions tested were 50, 10, 5, 1, 0.5 and 0.1 µg/ml. It was determined that 5 µg/ml of the LMWF generated by pectinase digestion exhibited the strongest inhibitory effect on osteoclasts (Fig. 1A). Based on these results, subsequent experiments were performed using 5 µg/ml of the LMWF generated by pectinase digestion.

The RAW264.7 murine macrophage-like cell line was purchased from the Chinese National Platform of Experimental Cell Resources for Sci-Tech, Institute of Basic Medicine, Chinese Academy of Medical Sciences (Peking Union Medical College, Beijing, China). RAW264.7 cells were grown in high glucose Dulbecco's modified Eagle's medium (DMEM; ME100202P1; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% heat-inactivated fetal bovine serum (12676011; Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin, at 37°C in a humidified atmosphere of 95% air and 5% CO₂. The cells were cultured on sterile coverslips or osteologic slides, which were placed into 6-well culture dishes. Cells were inoculated at a density of 2×104/ml. The medium was changed to high-glucose DMEM (ME100202P1; Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.), with 100 ng/ml RANKL (Murine sRANK-Ligand; 31501; PeproTech, Inc., Rocky Hill, NJ, USA) and 30 ng/ml M-CSF (Murine M-CSF, 31502; PeproTech, Inc.) to induce differentiation 4 h after cell attachment. Cells were incubated at 37°C, in a 5% CO₂ incubator with high humidity. The medium was changed every 2 days. After 6 days, cells were stained for tartrate-resistant acid phosphatase (TRAP) using the TRAP staining kit (Sigma-Aldrich; Merck Millipore) and the experiment was conducted according to the manufacturer's protocol. A light microscope (Olympus Corporation, Tokyo, Japan) was used to observe the number of TRAP-positive multinucleated cells (≥3 nuclei). To visualize bone resorption pits, osteologic slides were fixed in 2.5% glutaraldehyde, washed by sonication in phosphate-buffered saline (PBS), dehydrated in a graded, ascending series of ethanol, sputter-coated with gold and palladium. Images were captured using a scanning electron microscope.

To assess the effects of LMWF on osteoclast differentiation, RAW264.7 cells (2×104/ml) were cultured in two groups; A control group and an experimental, LMWF-treated group. Differentiation was induced as aforementioned, with the exception that the experimental group was cultured in the presence of 5 µg/ml LMWF in addition to RANKL and M-CSF. Cells were stained for TRAP and the number of TRAP-positive multinucleated cells (≥3 nuclei) was scored after 6 days. Scoring was performed at a magnification ×100 and 10 fields were randomly selected and scored, with the average calculated. Only cells that appeared pink or rose-red and contained ≥3 nuclei were counted, as previously described (19). The experiment was repeated three times. For the bone resorption pit assay, cells were cultured in as aforementioned, with the exception that they were cultured on bovine cortical bone slices made in-house. Cells were cultured on 10 bone slices in each group. Within each bone slice, 10 non-overlapping areas were randomly selected and analyzed using Image-Pro Plus version 6.0 image analysis software (Media Cybernetics, Inc., Rockville, MD, USA) to evaluate bone resorption areas.

Two groups of cells, as aforementioned, were cultured for 6 days. Subsequently, the cells (1×106/ml) were collected, washed with PBS, and fixed with 75% ethanol for 5 min at room temperature. Fixed cells were resuspended in 1 ml propidium iodide/Triton X-100 staining solution (Sigma-Aldrich; Merck Millipore) containing 0.2 mg RNase A and were incubated at 37°C for 15 min. Stained cells were subjected to flow cytometry using a Guava Flow Cytometer (model, EPICS® Merck Millipore) and Expo 32 software (Beckman Coulter, Inc., Brea, CA, USA) to calculate proliferation index (PI). PI = (S+G₂/M)/(G₀/G₁+S+G₂/M).

After 6 days of induction, cells were harvested for RNA extraction. Total RNA was isolated using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) The quality of RNA was assessed by agarose gel electrophoretic analysis. RANK, TRAP, matrix metallopeptidase-9 (MMP-9), NFATc1 and osteoclast-associated immunoglobulin-like receptor (OSCAR) expression levels were detected by qPCR using the Absolute™ Fast qPCR Master Mixes kit (Guangzhou Funeng Gene Co., Ltd., Guangzhou, China). The thermocycler conditions used were as follows: 96°C pre-denaturation for ~5 min; 94°C denaturation for ~1 min; 58°C annealing for ~45 sec; 72°C extension for ~2 min, for a total of 35 cycles. cDNA was reverse-transcribed using the All-in-One™ First-Strand cDNA Synthesis kit (GeneCopoeia, Inc., Rockville, MD, USA) from the isolated RNA and GAPDH expression was used as an internal control. Primers were designed by Fulengen Co., Ltd. (Guangzhou, China). The primer sequences were as follows: GAPDH, forward (F) 5′-CGGAGTCAACGGATTTGGTCGTAT-3′, reverse (R) 5′-AGCCTTCTCCATGGTGGTGAAGAC-3′, fragment amplified, 96 bp; NFATc1, F 5′-TCCAAAGTCATTTTCGTGGA-3′, R 5′-CTTTGCTTCCATCTCCCAGA-3′; MMP-9, F 5′-ACGACATAGACGGCATCCA-3′, R 5′-GCTGTGGTTCAGTTGTGGTG-3′; TRAP, F 5′-AAATCACTCTTTAAGACCAG-3′, R 5′-TTATTGAATAGCAGTGACAG-3′; RANK, F 5′-CCAGGGGACAACGGAATCA-3′, R 5′-GGCCGGTCCGTGTACTCATC-3′; and OSCAR, F 5′-TGATTGGCACAGCAGGAG-3′ and R 5′-AAGGCACAGGAAGGAAATAGAG-3′. RT and qPCR were performed per manufacturer's protocol. Amplification and melting curves were evaluated for each qPCR analysis. Gene expression levels were based on levels of the initial template, and were calculated using the 2-∆∆Cq method (20).

A total of 60 healthy 6-month-old female Sprague-Dawley (SD) rats weighing 220–240 g were provided by Tianjin Hospital Center for Experimental Animals (Tianjin, China; approval no. SCXK 2007004). All experimental procedures were approved by the Institutional Animal Care and Use Committee at the Tianjin Hospital Ethics Committee of China (Tianjin, China). The methods were carried out in accordance with the approved guidelines. Animals were housed at a room temperature of 21–24°C, in a lighted room between 7:00 a. m. and 7:00 p.m.; animals were allowed to engage in physical activity and received ad libitum access to food and water.

A total of 60 female SD rats were randomly assigned to the following treatment groups (n=20/group): i) Ovariectomized group (OVX); ii) sham surgery group (SHAM); and iii) LMWF group (LMWF dose: 5 mg/kg). SHAM group rats were subjected to a sham surgery (only an incision in the back was made) without ovary removal. The other two groups were subjected to bilateral ovary-removal surgery through an incision in the back following anesthesia with 10% chloral hydrate (0.3 ml/100 g). Saline (for the OVX and SHAM groups) or LMWF (for the LMWF group) was administered by intraperitoneal injection 5 days after surgery. Rats were euthanized with 10% chloral hydrate (1 ml/100 g) at 4 or 8 weeks following administration of treatment. A total of 10 rats from each group were euthanized at each time-point.

The bone mineral density (BMD) of the right femur of the rats (n=5 per group at each time-point) was measured using LUNAR DPXIQ bone densitometer (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA). Bones were then subjected to a micro-computed tomography (CT) scan (Bruker microCT, Antwerp, Belgium). Scanning was performed at a setting of 21-µm resolution, 360° rotation with an incremental rotation of 0.4°, voltage of 80 kV, current of 80 A and 2.960 msec exposure time. Images were captured at an average gain of 4, pixel 1×1. For image normalization, the images were scanned in black and white and the grids were standardized. Trabecular bone volume fraction (BV/TV, %), trabecular thickness (Tb.Th.) and trabecular number (Tb.N.) of the distal end of the femur were determined for evaluation.

For the 4-week and 8-week time-points, rats were injected with an intramuscular injection of tetracycline (25 mg/kg) at 10 days or 3 days prior to euthanasia, as previously described (21). Following euthanasia, the right femurs were collected and fixed in ethanol for 1 week. The distal 1/3 of each femur (n=5 per group at each time-point) was dehydrated and embedded in semi-butyl methacrylate polymerization/methyl ester. Using a heavy-duty microtome, 7 µm undecalcified bone sections were cut. The sections were washed in distilled water and treated with 3% sodium thiosulphate for 5 min, and then washed in water. von Kossa and Giemsa staining were performed for 5 min, respectively, and sections were the mounted. Images were captured using a light microscope and analyzed using Image-Pro Plus version 6.0 software.

Rats were anesthetized with 10% chloral hydrate (0.3 ml/100 g) and serum samples (1 ml) were collected by cardiac puncture 8 weeks after the administration of LMWF or saline control. Samples were centrifuged at 300 × g for 10 min at room temperature. Serum procollagen type I N propeptide (PINP) and C-terminal telopeptide-1 (CTX-1) levels were determined using corresponding ELISA kits (Ra PINP kit; cat. no. 20131010 and Ra CTX-1 kit; cat. no. 20130311; Cusabio Biotech Co., Ltd., Baltimore, MD, USA).

The mechanical strength of the shaft and neck of the left femur (n=10/group) was determined as previously described by Ma and Fu (22). All testing was performed on a Bose ElectroForce3200 electromagnetic test instrument (Bose Corporation, Framingham, MA, USA). The midshaft femoral strength was tested using a three-point bending test and the femoral neck strength was tested by a compression test (n=10).

Data are presented as the mean ± standard deviation. Statistical analysis was conducted using SPSS 16.0 software (SPSS, Inc., Chicago, IL, USA). All experimental data were analyzed using an one way analysis of variance and multiple comparisons were performed using Duncan's test. P<0.05 was considered to indicate a statistically significant difference.

As presented in Fig. 1B, 6 days after RANKL induction of RAW264.7 cells, the number of cells that stained positive for TRAP in the LMWF group was significantly lower compared with the control group (P<0.05), indicating that more osteoclasts were generated when LMWF was absent. In addition, TRAP-positive cells from the LMWF group had fewer nuclei in general (P<0.05; data not shown). The bone resorption assay revealed that there were significantly more bone resorption pits in the control group compared with the LMWF group (P<0.05; Fig. 1C).

FACS analysis revealed that the PI of cells from the LMWF group was significantly lower compared with cells from the control group (26.2±3.62 vs. 37.56±3.08, respectively; P<0.05; data not shown), thus suggesting that LMWF inhibited osteoclast proliferation.

RT-qPCR was performed to assess expression levels of the genes encoding TRAP, RANK, NFATc1, OSCAR and MMP-9. As presented in Fig. 2, LMWF significantly inhibited the expression levels of osteoclast marker genes TRAP (P<0.05), NFATc1 (P<0.01) and OSCAR (P<0.05) when compared with the control group. However, LMWF had no effect on RANK gene expression (P>0.05). LMWF also exhibited an inhibitory effect on MMP-9 expression (P<0.05).

As presented in Fig. 3A, 4 weeks after the ovariectomy the distal femur of the OVX group exhibited trabecular bone loss, which progressively worsened 8 weeks after surgery when compared with the SHAM group. As presented in Fig. 3B, the BMD of rats from the OVX group was significantly lower for both time-points investigated compared with that of rats from the SHAM group (P<0.05), thus suggesting that ovariectomy may lead to bone loss. The ovariectomized rats receiving LMWF had a significantly higher BMD compared with the OVX group (P<0.05; Fig. 3B), indicating that LMWF may minimize bone loss. In addition, the BMD increase at the 8-week time-point was higher than that at the 4-week time-point; however, still lower when compared with the SHAM group.

Quantification of micro-CT scan results are presented in Fig. 3C-E. Compared with the OVX group, the LMWF group demonstrated a significantly higher BV/TV, Tb.Th. and Tb.N at 8 weeks (P<0.05 for all measurements), whereas these measurement were lower compared with in the SHAM group. In addition, the rats of the LMWF group demonstrated greater Tb.N in the distal femur compared with the OVX group.

Fig. 4 presents the quantitative histomorphometric analyses of the undecalcified histological sections stained with (A) von Kossa and (B) Giemsa. The OVX group exhibited higher tetracycline labeling (Fig. 4C), thicker osteoid seams, and a higher rate of mineralization on osteoblast surfaces and osteoclast surfaces when compared with the SHAM group (Fig. 4D; P<0.05), which was consistent with the high bone turnover usually observed in osteoporosis. LMWF administration reduced bone turnover rate in OVX rats; however, the surface area for bone formation and the mineralization time was increased compared with the OVX group (Fig. 4D; P<0.05).

As presented in Fig. 5A, the maximal strength of the femur as assessed by the three-point bending test and compression test was significantly lower in the OVX group when compared with the SHAM group rats (P<0.05). The maximum load of femoral neck in the OVX group was significantly reduced 8 weeks after surgery when compared with 4 weeks after surgery (P<0.05). Mechanical strength of femoral neck and mid-shaft femoral in OVX rats was lower compared with the SHAM group (P<0.05; Fig. 5A). The LMWF group demonstrated higher femoral mechanical strength of femoral neck and mid-shaft femoral in rats 4 and 8 weeks after ovariectomy when compared with the OVX group (P<0.05; Fig. 5A).

The OVX group had significantly increased levels of CTX-1 and PINP in serum compared with the SHAM group rats at 4 and 8 weeks after the operation. As presented in Fig. 5B LMWF treatment significantly decreased the levels of serum PINP (P<0.01 for both time-points) and CTX-1 (P<0.01 at 4 weeks and P<0.05 at 8 weeks) compared with the OVX group rats.