All experiments were performed in strict accordance with the appropriate guidelines (30) and were approved by the Animal Care and Use Committee of the Medical School of Nanjing University (Nanjing, China).

For in vivo experiments, male C3H/HeN mice (weight, 20–25 g; age 4–6 weeks, n=141) were purchased from Vital River Experimental Animal Corporation of Beijing. In total, 6 mice were housed in one cage under a 12 h light/dark cycle at 20°C, with a relative humidity of 55% and with free access to water and food.

For in vitro experiments, 14-day pregnant Sprague-Dawley rats were used to obtain the fetuses and collected the primary neuronal cells from the fetuses. Sprague-Dawley rats in the 14th day of pregnancy (weight, 300–350 g; age 6 weeks, n=3) were purchased from Qing Long Shan Dong Wu Fan Zhi Chang (Jiangsu, China, http://www.njqlsdwc.com). Rats were housed one cage per rat under a 12 h light/dark cycle at 20°C, with a relative humidity of 55% and with free access to water and food.

A total of 64 mice were randomly divided into eight groups (n=8): Control group, sham group, and tumor group, pain behavioral tests were performed on the day before (baseline) and on day 4, 7,10, 14, 21 and 28 after operation. sham + vehicle group, tumor + vehicle group, tumor + JWH015 (1 µg) group, tumor + JWH015 (2 µg) group, and tumor + JWH015 (1 µg) + AM630 (2 µg) group, pain behavioral tests were performed on the day before (baseline) and at 4, 8,12, 24, 48 and 72 h after injection.

A total of 20 mice were randomly divided into two groups: Sham group (n=6), and tumor group (n=24). 14 days after operation, mice in sham group were sacrificed and the lumbar spinal cord was collected for western blotting (n=3) and immunofluorescence labeling (n=3). Mice in tumor group were sacrificed on day 0 (n=3), 4 (n=3), 7 (n=3), 10 (n=3), 14 (n=3), 21 (n=3), and 28 (n=3) for western blotting. On day 14 after operation, the mice in tumor group were sacrificed for immunofluorescence labeling (n=3).

A total of 36 mice were randomly divided into eight groups: Sham + vehicle group (n=6), tumor + vehicle group (n=6), sham + Baf-A1 group (n=3), tumor + Baf-A1 group, tumor + JWH015 (1 µg) group (n=3), tumor + JWH015 (2 µg) group (n=6), tumor + JWH015 (1 µg) + AM630 (2 µg) group (n=6), and tumor + JWH015 + Baf-A1 group (n=3). Mice were sacrificed at 12 h after injection in each group for western blotting (n=3). In sham + vehicle group, tumor + vehicle group, tumor + JWH015 (2 µg) group, tumor + JWH015 (1 µg) + AM630 (2 µg) group, another 3 mice were sacrificed at 12 h after injection for immunofluorescence labeling (n=3).

A total of 21 mice in tumor + JWH015 (2 µg) group were randomly sacrificed at 0 (n=3), 4 (n=3), 8 (n=3), 12 (n=3), 24 (n=3), 48 (n=3), and 72 (n=3) h after injection for western blotting.

NCTC 2472 osteolytic sarcoma cells (cat. no. 2087787; American Type Culture Collection) were cultured in NCTC 135 medium (Sigma-Aldrich; Merck KGaA) containing 10% horse serum (Gibco; Thermo Fisher Scientific, Inc.) and maintained in a 5% CO2 atmosphere at 37°C (Thermo Forma; Thermo Fisher Scientific, Inc.).

Sprague-Dawley rats in the 14th day of pregnancy were deeply anesthetized with 2–3% isofluorane, sacrificed by cervical dislocation and the fetuses were quickly removed on embryonic day 14. The meninges and blood vessels were removed from the fetal cerebral cortices under a dissecting microscope. After cutting into 1 mm³ pieces, cortical tissues were digested in 0.25% trypsin at 37°C for 10 min. Supernatants were passed through a 70 µm cell strainer (Falcon; Thermo Fisher Scientific, Inc.) and centrifuged for 5 min at 168 × g at 37°C. Cells were diluted in Neurobasal medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 1% B27 (Gibco; Thermo Fisher Scientific, Inc.), 2 mM glutamine and 10 µl/ml penicillin/streptomycin and plated onto poly-L-lysine (Sigma-Aldrich; Merck KGaA) coated 6-well plates at a density of 1×10⁶ cells/cm². Cells were cultured at 37°C in a humidified incubator containing 5% CO₂. The culture medium was changed 3 days after plating and cells were allowed to grow for 7 days before using in subsequent experiments.

The BCP model was constructed as described by Schwei et al (31). Mice were anesthetized with sodium pentobarbital (50 mg/kg; i.p.) and an incision was made in the skin on the right leg and the right joint was exposed. Then a hole was drilled in the femur-plateau. Subsequently, 20 µl α-minimum essential medium (Thermo Fisher Scientific, Inc.) containing 2×10⁵ NCTC 2472 osteolytic sarcoma cells was injected into the intramedullary space of the right femur. Mice in the Sham group were injected with the isodose medium without any cells. The drilled hole was sealed with bone wax, and the wound was closed with 4-0 silk sutures (Ethicon, Inc.). Mice recovered from anesthesia on a heated blanket.

For in vivo experiments, drugs were prepared and administered as previously described (29,32,33). The selective CB2 agonist JWH015 (Sigma-Aldrich; Merck KGaA) was dissolved in 5% DMSO corresponding to a dose of 1 µg/5 µl (50 µg/kg) or 2 µg/5 µl (100 µg/kg). The CB2-selective antagonist AM630 (Sigma-Aldrich; Merck KGaA) was dissolved in 5% DMSO corresponding to a dose of 2 µg/5 µl (100 µg/kg). Bafilomycin A1 (Baf-A1), an inhibitor of autophagosome and lysosome fusion, was dissolved in 5% DMSO corresponding to a dose of 10 nM. To avoid systemic effects on tumor cells, intrathecal administration was selected (34). Drugs were intrathecally administered at a volume of 5 µl at day 14 after the inoculation with tumor cells. To antagonize the activation of the CB2 receptor, AM630 was injected 30 min before JWH015.

For in vitro experiments, JWH015 and AM630 were dissolved in culture medium corresponding to a dose of 1 µM. Lipopolysaccharide (LPS) was dissolved in culture medium corresponding to a dose of 100 nM.

Mechanical allodynia and spontaneous pain in mice were tested prior to operation (day 0) as well as 4, 7, 10, 14, 21 and 28 days after operation in each group, and 0 (baseline), 4, 8, 12, 24, 48 and 72 h after the administration of JWH015, AM630 and vehicles. The experimenters who performed all behavioral tests were blinded to the groups.

PWMT in the right hind paw was measured using von Frey filaments (0.16, 0.4, 0.6, 1.0, 1.4 and 2.0 g; Stoelting, USA) and the ‘up-down’ method as previously described (33). Mice were placed in transparent plexiglass compartments with a wire mesh bottom and allowed to acclimate for 30 min. von Frey filaments were stuck to the plantar surface, and the lowest filament stimulus strength that resulted in the paw flinching or withdrawing was regarded as the PWMT.

Mice were placed in transparent plexiglass compartments with a wire mesh bottom and allowed to acclimatize for 30 min. Subsequently, the NSF of the right hind paw in 2 min was counted; each mouse was tested five times.

Mice were anesthetized with pentobarbital (50 mg/kg, i.p.) and sacrificed by cervical dislocation on day 0, 4, 7, 10, 14, 21 and 28 after operation and 0, 4, 8, 12, 24, 48 and 72 h after intrathecal administration. The L3-L5 segments of the spinal cord were removed and stored at −80°C for further study. In addition, the primary neuronal cells were collected at 0, 3, 6, 9, 12 and 24 h after LPS-stimulation, and 12 h after JWH015-treatment. Samples were homogenized in RIPA Lysis Buffer (10 µl/mg for tissue; Beyotime Institute of Biotechnology) with phenylmethyl sulfonyl fluoride and incubated on ice for 30 min, followed by centrifugation at 241 × g at 4°C for 20 min. The supernatant of each sample was collected. Protein concentrations were determined using a BCA Protein Assay kit. Each sample of 50 µg protein was subjected to 10% SDS-PAGE and then transferred onto a PVDF membrane. The membrane was blocked with 5% BSA (Gibco; Thermo Fisher Scientific, Inc.) at room temperature for 1 h. The membranes were incubated with primary antibodies against IL-1β (1:1,000; Abcam; cat. no. ab2105), IL-6 (1:1,000; Abcam; cat. no. ab6672), LC3B (1:1,000; Cell Signaling Technology, Inc.; cat. no. 3868), p62 (1:1,000; Abcam; cat. no. ab91526) and β-actin (1:4,000; Abcam; cat. no. ab8227). The blots were subsequently incubated with a horseradish peroxidase conjugated goat anti-rabbit secondary antibody (1:10,000; EMD Millipore; cat. no. AP132P) and developed in ECL solution (Tanon Science and Technology Co., Ltd.). Images were captured using a cooled CCD system (Tanon Science and Technology Co., Ltd.) and quantified using ImageJ v1.8.0 (National Institutes of Health).

Following the administration of general anesthesia (pentobarbital, 50 mg/kg; i.p.), mice were transcardially perfused with normal saline and 4% paraformaldehyde at day 14 after operation and 12 h after JWH015 (2 µg) administration. Lumbosacral enlargements were removed and fixed in 4% paraformaldehyde for 6 h, and then dehydrated in 30% sucrose for 48–72 h at 4°C. After they were frozen in optimal cutting temperature compound, tissues were cut into 20 µm sections with a freezing microtome. Sections were placed in PBS and sequentially blocked with 10% goat serum (Gibco; Thermo Fisher Scientific, Inc.) containing 0.3% Triton (Tanon Science and Technology Co., Ltd.) for 2 h at room temperature. Sections were then incubated with primary antibodies for glial fibrillary acidic protein (GFAP; mouse, 1:100, Cell Signaling Technology, Inc. #80788), ionized calcium-binding adaptor molecule 1 (Iba1; rabbit, 1:300, Wako, Japan), LC3B (rabbit; 1:100; Cell Signaling Technology, Inc.; cat. no. 3868), and neuronal nuclei antigen (NeuN; mouse; 1:1,000; Abcam; cat. no. ab104224) separately overnight at 4°C. After washing with PBS, sections were incubated with Alexa 488-conjugated goat anti-rabbit (1:3,000, Thermo Fisher Scientific, Inc.; cat. no. R37116) or Alexa 594-conjugated goat anti-mouse (1:3,000; Thermo Fisher Scientific, Inc.; cat. no. R37121) secondary antibodies. The sections were mounted on glass slides, air-dried and incubated with DAPI (Abcam) for 5 min at room temperature for nuclear staining. Images were captured using a laser-scanning confocal microscope (Olympus Corporation) and the staining density was analyzed using ImageJ (National Institutes of Health).

Data are presented as the mean ± standard deviation. Results from the behavioral study were analyzed using repeated measurements ANOVA followed by Bonferroni test post-hoc test to assess differences at each time point among and with. Western blotting results were analyzed using one-way ANOVA followed by Bonferroni test for between-group comparisons. Statistical analyses were performed using SPSS 22.0 (IBM Corporation); GraphPad Prism Version 7 (GraphPad Software, Inc.) was used to plot graphs. P<0.05 was considered to indicate a statistically significant difference.

PWMT and NSF of the right hind paw of mice were tested to monitor the progression of BCP generated by injecting NCTC 2472 osteolytic sarcoma cells into the femur. No significant differences in PWMT and NSF were observed between groups at baseline. PWMT decreased slightly (1.15±0.207 g in the Sham group; 1.30±0.185 g in the tumor group; Fig. 1A) and NSF increased (3.500±0.756 in the Sham group; 2.875±0.641 in the tumor group; Fig. 1B) on day 4 in the Sham and Tumor groups compared with the values at baseline (all P<0.05). The PWMT and NSF values recovered to near baseline level in the Sham group from day 7. In the Tumor group, PWMT recovered on day 7 and decreased starting on day 10 (0.9±0.28 g on day 10; 0.52±0.21 g on day 14; 0.34±0.11 g on days 21 and 28; P<0.05; Fig. 1A) compared with the baseline. NSF in the Tumor group remained significantly higher compared with the baseline value at days 10–28 (5.875±1.12599 on day 10; 9.5±0.92582 on day 14; 11±1.06904 on day 21 and 12.5±0.92582 on day 28; P<0.05; Fig. 1B). These results confirmed the successful establishment of the bone cancer pain model.

Activation of CB2 by intrathecal administration of JWH015 on day 14 significantly improved the pain behaviors. There were no significant differences were identified in PWMT and NSF in the Sham group following vehicle treatment. JWH015-treated mice exhibited a large increase in PWMT and a decrease in NSF in a dose-dependent manner (Fig. 1C and D, respectively). PWMT increased and NSF decreased from 8 to 48 h after administration in the Tumor + JWH015 (1 µg) group (P<0.05) and the Tumor + JWH015 (2 µg) group (P<0.05) compared with the respective baseline values. The analgesic effect of JWH015 peaked at 12 h after injection, with an increase in PWMT to 1.85±0.278 g in the Tumor + JWH015 (2 µg) group and 1.3±0.185 g in the Tumor+JWH015 (1 µg) group. However, the pain-relief effect of JWH015 was completely prevented by pretreatment with the CB2 antagonist AM630 (Fig. 1C and D).

There are two major autophagic marker proteins, LC3B and p62. LC3B-I is converted to LC3B-II during the formation of autophagosomes (35), whereas p62 is degraded by the autophagosome-lysosome pathway and represents the degradation level of autophagosomes (36). To determine whether autophagy flux was impaired in the spinal cord during BCP, western blotting was performed. The results indicated that LC3B-II the ratio of LC3B-II/LC3B-I were significantly increased in BCP mice at 14, 21 and 28 days after operation compared with Sham group (Fig. 2A and B; P<0.05). In addition, the expression of p62 was significantly increased from day 10 to 28 (Fig. 2C; P<0.05).

Baf-A1 suppresses autophagosome-lysosome fusion and results in the accumulation of autophagosomes. As shown in Fig. 2D-F, intrathecal injection of 10 nM Baf-A1 increased the ratio of LC3B-II/LC3B-I and the expression of p62 in the Sham mice compared with the values in the Sham + vehicle mice (P<0.05). These results indicated that autophagy flux was impaired in BCP model mice.

The cellular localization of LC3B expression was determined using double immunofluorescence staining. The spinal tissues were collected on day 14 from the Tumor groups. The results demonstrated substantial LC3B expression in the spinal dorsal horn, which was mostly colocalized with NeuN, a neuronal marker (Fig. 3A). The expression of LC3B in astrocytes and microglia was also examined. No colocalization with GFAP (an astrocyte marker; Fig. 3B) and Iba1 (a microglial marker; Fig. 3C) in the spinal cord was found. These results indicated that the autophagy flux in the dorsal horn neurons was impaired in BCP.

Based on the protective role of autophagy in pain and the relationship between autophagy and the CB2 in several diseases (19–24), the relationship between the CB2 and autophagy was involved in BCP was investigated. There were significant increase of the ratio of LC3B-II/LC3B-I and the expression of p62 in the Tumor + vehicle group compared with the Sham + vehicle group (P<0.05). Intrathecal administration of JWH015 decreased the ratio of LC3B-II/LC3B-I in the Tumor + JWH015 (1 µg) and Tumor + JWH015 (2 µg) groups, as well as decreased the expression of p62 in the tumor + JWH015 (2 µg) group, compared with the values in the Tumor + vehicle group (Fig. 4A-C; P<0.05). Pretreatment with the CB2 antagonist AM630 reversed the increased autophagy flux in the JWH015-treated mice. The protein expression levels of LC3B and p62 were also examined at different time points following JWH015 (2 µg) injection; the ratio of LC3B-II/LC3B-I and the expression of p62 decreased from 8 to 72 h after injection (Fig. 4D-F; P<0.05). In addition, the ratio of LC3B-II/LC3B-I and the expression of p62 increased in the Tumor + JWH015 (2 µg) + Baf-A1 (10 nM) group compared with the values in the Tumor + JWH015 (2 µg) group (Fig. 4G-I; P<0.05). Collectively, these results suggested that JWH015 ameliorated the impaired autophagy flux in BCP.

Although the results aforementioned demonstrated the involvement of CB2-mediated autophagy flux in BCP, the mechanism by which CB2 affected autophagy was not clarified. Accumulating evidence has indicated the role of inflammation as the bridge between the CB2 and autophagy (21,22). Thus, whether the amelioration of autophagy flux by JWH015 was mediated by the modulation of inflammation was determined.

Glial cell-derived pro-inflammatory mediators, such as IL-1β and IL-6, have been reported to play key roles in the pathophysiology of pain (29). Increases in the staining density of GFAP (P<0.05) and Iba1 (P<0.05) were observed on day 14 after operation in BCP mice compared with the levels in the Sham mice (Fig. 5A). Western blotting revealed the significantly increased expression levels of IL-1β on days 4, 14, 21 and 28 (Fig. 6A and B; P<0.05) and of IL-6 on day 14 (Fig. 6A and B; P<0.05) in the spinal cord in BCP mice compared with the levels in the Sham mice.

Intrathecal administration of JWH015 at a dosage of 2 µg decreased the staining density of GFAP and Iba1 at 12 h after treatment (Fig. 5B; P<0.05) compared with the values in the Tumor + vehicle group. Compared with the values in the Tumor + vehicle group, intrathecal injection of 1 and 2 µg JWH015 significantly lowered the protein expression levels of IL-1β and IL-6 (Fig. 6C and D; P<0.05). The expression of IL-1β and IL-6 decreased between 8 and 72 h after JWH015 (2 µg) injection (Fig. 6E and F, P<0.05). However, the suppressive effect of JWH015 was completely prevented by pretreatment with AM630 (Fig. 6C and D; P>0.05).

In summary, the activation of CB2 by the intrathecal administration of JWH015 inhibited the activation of glial cells, and downregulated glial cell-derived IL-1β and IL-6, and this may be the underlying mechanism by which autophagy flux was ameliorated.

Previous studies have reported the increased expression of IL-1β and IL-6 following LPS-stimulation in various cells (37,38). Therefore, LPS-stimulated primary neurons were used as an in vitro model to further explore the role of IL-1β and IL-6 in the activation of autophagy by JWH015. Primary neurons were stimulated with LPS (100 nM) for 0, 3, 6, 9, 12 or 24 h. Western blotting revealed significantly increased protein expression levels of IL-1β and IL-6 in primary neurons at 9, 12 and 24 h after LPS stimulation, compared with untreated Control cells (Fig. 7A and B; P<0.05), which indicated an inflammatory environment induced by LPS. IL-1β and IL-6 were upregulated over time following LPS stimulation, along with increases in the ratio of LC3B-II/LC3B-I and expression levels of p62 at 9, 12 and 24 h after stimulation (Fig. 7C and D; P<0.05), which indicated the impairment of autophagy flux induced by the LPS-mediated production of IL-1β and IL-6 in primary neurons. Treatment with JWH015 at a dosage of 1 µM for 12 h in LPS stimulated-primary neurons significantly reduced IL-1β and IL-6 expression levels (Fig. 7E and F; P<0.05). In addition, the ratio of LC3B-II/LC3B-I and the expression of p62 decreased in the JWH015-treated group (Fig. 7G and H; P<0.05). These effects could be prevented by pretreatment with AM630 as the expression of IL-1β and IL-6, the ratio of LC3B-II/LC3B-I, and the expression of p62 was increased in the LPS + JWH015 + AM630 group compared with the LPS + JWH015 group (P<0.05).

In summary, the results of the in vitro experiment demonstrated that the impairment of autophagy flux was induced by the production of IL-1β and IL-6 in LPS-stimulated primary neurons. JWH015 may ameliorate autophagy by the downregulation of the inflammatory mediators IL-1β and IL-6.