Dulbecco’s modified Eagle’s medium (DMEM) and fetal bovine serum (FBS) were obtained from Gibco, USA. PD168393 was purchased from Calbiochem, and HRG1β (ab50277) was obtained from Abcam, UK. Rabbit anti-ErbB2 (ab2428) and anti-p-ErbB2 (ab47755) were bought from Abcam. Mouse anti-NRG1 (SC-57384), rabbit anti-p-AKT-1 and rabbit anti-p-p38MAPK were purchased from Santa Cruz, USA. Anti-GAPDH was obtained from Sigma, USA. PVDF membrane and ECL Western Blotting Substrates were products of Millipore, USA and Pierce, USA, respectively. SYBR PrimeScript™ RT-PCR kit was obtained from Takara, Japan. Designed PCR primers were synthesized by Takara Biotechnology (Dalian) Co., Ltd, China.

The Walker 256 mammary gland carcinoma cell line was purchased from Cancer Institute and Hospital, Chinese Academy of Medical Sciences, Beijing, China. Cells were maintained in DMEM with 10% fetal bovine serum and penicillin/streptomycin (100 U/ml). The harvested cells were suspended at a concentration of 2×10⁷ cells/ml. Female Sprague-Dawley (SD) rats, weighing 80–100 g, were provided by Experimental Animal Center, Shengjing Hospital of China Medical University, Shenyang, China. Each rat received intra-abdominal injection of 0.5 ml of the cell suspension. Ascites formed 6–7 days following tumor cell injection. Thereafter, 5 to 10 ml of ascites were collected and cell were harvested by centrifugation, then suspended in phosphate buffer saline (PBS) at a density of 1×10⁷ cells/ml. Same amount of cells were taken as a negative control by heat inactive treatment in a 100°C water bath for 10 min.

Female SD rats as mentioned above, weighing 200–220 g, were routinely housed in light (12 h dark/12 h light) and temperature-controlled (22–24°C) rooms and had free access to food and water. Experimental animal use was approved by the Experimental Ethics Committee of Shengjing Hospital of China Medical University. Animal experiments were carried out in accordance with the Guidelines laid down by the NIH in the US regarding the care and use of animals for experimental procedures. The rat model of CIBP was established as previously described (20). In brief, rats were anesthetized by intraperitoneal injection of 10% chloral hydrate (300 mg/kg). The right leg was shaved and the skin was disinfected with 70% (v/v) ethanol. A 1-cm incision was made in the skin, and the tibia was carefully exposed with minimal damage to blood vessels and nerve. The tibia was pierced using a 23-gauge needle. A total of 10 μl of cell suspension in PBS that contains 0.1 million (10⁵) Walker 256 mammary gland carcinoma cells were slowly injected into the right tibia cavity of each rat using a 25-μl microinjection syringe within 2 min. The injection site was sealed with bone wax, followed by irrigation with sterile normal saline.

Intrathecal catheterization was carried out as previously described (21). In brief, animals were anesthetized by 10% chloral hydrate (300 mg/kg, i.p.). A PE10 intrathecal catheter was inserted 2 mm cephalad into the rat lumbar subarachnoid space at the level of L3–L4 inter-vertebrae. The tip of the catheter was located near the lumbar enlargement of the spinal cord for intrathecal administration of the drugs. The catheter was tunneled subcutaneously and externalized through the skin in the neck region. Two days later, 20 μl lidocaine (20 g/l) was injected intrathecally to rats without impaired movement. If these animals showed lower limb paralysis within 30 sec of injection time, it was a successful sign for catheterization.

To verify the effect of ErbB2 signaling on the regulation of CIBP, PD168393, a structurally related irreversible-binding ErbB2 inhibitor (22), was used. Animals were randomly divided into four groups, including: sham operation (n=6 for each time point), CIBP (n=6 for each time point), CIBP + PD168393 treatment (n=12), and CIBP + vehicle treatment (n=12). In CIBP group, intact Walker 256 mammary gland carcinoma cells were injected into the tibia cavity of rat without any treatment, whereas in the sham operation group, the same amount of heat-inactivated cells (negative control) were injected. In the CIBP + PD168393 or CIBP + vehicle groups, 10 μg of PD168393 was dissolved in 5% DMSO (10 μl), or the same volume of DMSO was administered by intrathecal injection on the 6th day after pain model establishment, respectively. Drug was injected daily for 9 successive days.

To observe the effect of NRG1 on the induction of cancer-induced bone pain, animals were randomly divided into three groups, sham operation, HRG1β and HRG1β + PD168393. Six animals were used for each time point. In the sham operation group, 10 μl of normal saline was administered by intrathecal injection; while in the HRG1β group, 10 μl of exogenous NRG1 (HRG1β, 0.4 ng/μl) was administered by intrathecal injection. In the HRG1β + PD168393 group, 1 h post-injection of 10 μg PD168393, the same dose of HRG1β was injected. Drug injection was performed once a day for three successive days.

Thermal hyperalgesia was determined by measuring paw withdrawal latency (PWL) during heat stimulation, using a thermal stimulator manufactured by Biomedical Engineering Institute of Chinese Academy of Medical Science (BME-410A). As described previously (23,24), rats were placed in a container (22 cm × 12 cm × 12 cm; length × width × height) with a smooth glass floor. All rats were given 30 min to acclimate to the testing environment. A heat source was focused on a portion of the hindpaw, which was flush against the glass and a radiant thermal stimulus was delivered to that site. The stimulus shut off automatically when the hindpaw moved, or after 20 sec to avoid tissue damage. Thermal stimuli were delivered three times to each hindpaw at 10-min intervals. Baseline thermal withdrawal latency for each rat was measured one day prior to experiment.

Mechanical paw withdrawal thresholds (PWT) were measured with von Frey filaments using a modification of the up-and-down method (25). Rats were placed in transparency organic glass cages with wire mesh bases. A total of 30 min of time for the acclimation of testing environment was given. A series of von Frey filaments with bending pressures ranging from 0.6 to 26 g were used. Those von Frey hairs were pressed perpendicularly against the plantar surface of the hind paw and held for 3–4 sec. Each von Frey filament was applied 5 times with 15-sec intervals. A positive response was regarded in presence of sharp withdrawal of the paw, licking of the paw or flinching upon removal of the von Frey filament. Baseline PWT for each rat was measured one day prior to experiment.

To examine tibial destruction from the inoculated tumor, animals were radiographed at 6, 14 and 21 days following tumor cell inoculation. Animal right hind limbs were placed on X-ray film and exposed to an X-ray source for 1/20 sec at 40 kV (p) at day 12 post-inoculation. The radiographic images were visually quantified using the scale developed by Schwei et al (26).

Animals were euthanized under anesthesia three weeks following establishment of the CIBP model. The right tibia and surrounding tissues were carefully removed, fixed in 4% paraformaldehyde (PFA), decalcified in PFA solution containing 5% formic acid, embedded in paraffin and sectioned. Sections were deparaffinized by immersing the tissues in dimethylbenzene, followed by rehydration. Haematoxylin and eosin (H&E) staining was performed according to standard procedures.

Animals were sacrificed and the L4–L6 lumbar spinal cord segments were dissected. Total RNA was extracted using TRIzol Reagent according to the manufacturer’s protocol (Invitrogen, USA). cDNA was then transcribed using PrimeScript™ RT Enzyme Mix I, Oligo (dT) primer, Random 6 mers and 5xPrimeScript™Buffer (Takara). The sequences of specific primers are listed in Table I. PCR reactions with 25 μl volumes was carried out using a LightCycler Quantitative PCR instrument (Roche, USA) in the following conditions: a preheating cycle at 95°C for 30 sec, followed by 40 cycles at 95°C for 5 sec, specific annealing temperatures are indicated in Table I for 15 sec, and 72°C for 10 sec. The relative expression from amplified RNA samples was calculated using the 2^−ΔΔCT method (27).

Protein was extracted from L4–L6 lumbar spinal cord segments by homogenization with radio immunoprecipitation assay (RIPA) lysis buffer and protein concentration was measured using a bicinchoninic acid (BCA) Protein Assay reagent kit (Santa Cruz) according to the manufacturer’s instructions. Equal amounts of protein extracts were then separated by 6 or 10% sodium dodecyl sulfate-polyacryl-amide gel electrophoresis (SDS-PAGE), transferred to a PVDF membrane, then blocked with 5% (w/v) non-fat dry milk in tris buffered saline plus Tween-20 (TBS-T; 0.1% Tween-20), followed by overnight incubation with primary antibodies at 4°C. The applied primary antibodies were diluted with TBS-T as follows: anti-ErB2 (1:200), anti-p-ErB2 (1:200), anti-NRG1 (1:200), anti-p-Akt-1 (1:500), anti-p-p38MAPK (1:500), and anti-GAPDH (1:1,000). After washing with TBS-T, the membranes were incubated with HRP labeled secondary antibodies (1:5,000). The expression level of target proteins was determined by analysis using an Image Acquisition and Analysis System (GDS8000, UVP, Inc., USA). The densitometric values were semiquantified by Bio-Rad and Quantity One imaging analysis software. The housekeeping protein, GAPDH, was used as an internal control.

Statistical analysis was performed using SPSS 16.0 software. Data were presented as the mean ± standard deviation (SD). P-value <0.05 was considered to be significant. Data were analyzed using a one-way and repeated measures analysis of variance (ANOVA) followed by Bonferroni-corrected post-hoc tests.

Radiological and histological examinations were used to evaluate bone destruction in the rat model of CIBP. As shown by X-ray images, no obvious alternation of tibia structure was found at either the 6th, 14th, or 21st day after inoculation of heat-inactivated Walker 256 mammary gland carcinoma cells (sham) (Fig. 1A; bone destruction score: 0; n=6). However, 6 days following inoculation of Walker 256 tumor cells, defects within the proximal tibial epiphysis were detected. These defects became more prominent at the 14th day (bone destruction score: day 6, 1.66±0.52; day 14, 2.66±0.54; n=6 for each). Further deterioration was found at 21 days post-injection, with the occurrence of cortical destruction, tumor dissemination and soft tissue tumors (bone destruction score: 4.33±0.43; n=6). Histological examination showed that bone marrow spaces were infiltrated with tumor cells at the 21st day following tumor cell inoculation (Fig. 1B). Moreover, bone trabecula defects, loss of normal bone structure, cortical destruction, and soft tissue tumors were observed in the CIBP model group. Bone destruction was not observed in animals within the sham operation group. A progressive decrease in both PWL and PWT, corresponding to thermal and mechanical stimulation of the inoculated hind paw, occurred at the 6th day after tumor cell injection (Fig. 1C and D). Compared with the sham operation group, significant differences in PWT and PWL values in the CIBP group were found at days 6, 10, 14 and 21 following tumor cell inoculation (P<0.01).

Next, we examined the expression level of NRG1 and ErbB2 in the tissues of L4–L6 spinal cord space from rats inoculated with intact or heat-inactivated Walker 256 mammary gland carcinoma cells. Compared with sham operation control, the expression levels of mRNA and protein of NRG1, ErbB2 and p-ErbB2 were significantly elevated at day 6 after tumor cell inoculation (P<0.05) (Fig. 2). Time-dependent upregulation of NRG1, ErbB2 and p-ErbB2 expression was observed within three weeks following tumor cell injection. These observations suggest that sustained activation of NRG1 and ErbB2 signaling in the spinal dorsal tissues after intra-tibia inoculation with tumor cells might be involved in development of CIBP in the rat model.

To clarify the role of ErbB signaling pathway in CIBP, PD168393, a selective tyrosine kinase inhibitor of ErbB2 was used in this study. The inhibitor was administered daily, intrathecally, beginning 6 days after tumor cell inoculation. As shown in Fig. 3, after 4 days of treatment with the ErbB2 inhibitor, CIBP-induced thermal hyperalgesia and mechanical allodynia were significantly attenuated 10 days after tumor cell inoculation (P<0.01, compared to CIBP group). These protective effects were sustained until 14th day after tumor cell inoculation, afterwards gradually decreasing until day 21, likely due to drug withdrawal at day 15. The expression of NRG1 was not impacted by the ErbB2 inhibitor.

To confirm the inhibitory effect of PD168393 on ErbB2 signaling activity, we examined the CIBP-induced phosphorylation of ErbB2 in spinal cord tissues of rats. The results showed that the level of phosphor-ErbB2 was significantly reduced in spinal cord tissues of rats in CIBP + PD168393 group, compared with CIBP or CIBP + DMSO (P<0.01) (Fig. 4A). However, treatment with PD168393 had no effect on the expression levels of NRG1 mRNA and protein, as compared with CIBP group (Fig. 4B and C), suggesting NRG1 may act as an upstream regulator of ErbB2 signaling pathway.

As shown in Fig. 5A and B, activation of Akt-1 and P38MAPK was detected in rats with intratibia injection of tumor cells. The phosphorylation level of Akt-1 and P38MAPK gradually increased from the 6th day after tumor inoculation, and reached a peak at the 14th day, then decreased until day 21. After 9 days of treatment with PD168393, the mRNA expressions of Akt-1 and P38MAPK were greatly reduced and the protein levels of p-Akt-1 and p-P38MAPK were significantly reduced in spinal cord tissues of rats, as compared with those in CIBP group (P<0.01) (Fig. 5C–E). However, no significant difference in levels of p-Akt-1 and p-P38MAPK was found between CIBP group and CIBP + DMSO group (Fig. 5D and E).

To validate the effect of NRG1 on pain, we observed the pain-like behavior of rats with intrathecal injection of exogenous NRG1. We found that intrathecal administration of exogenous NRG1 (HRG1β) could trigger thermal hyperalgesia and mechanical allodynia in rats immediately after injection. In our observation, the values of PWL and PWT were significantly reduced after an administration of HRG1β in a time-dependent pattern (P<0.01, compared with sham group) (Fig. 6). However, co-treatment with HRG1β and PD168393 relieved the exogenous NRG1-induced thermal hyperalgesia and mechanical allodynia (P<0.01 compared to treatment with HRG1β alone) (Fig. 6).

Our study also found that administration of exogenous NRG1 induced an increase in mRNA and protein expression of ErbB2, as well as the level of p-ErbB2 (Fig. 7A, C and D). Interestingly, the activation of ErbB2 could be inhibited by co-administration of PD168393 (Fig. 7D). Nevertheless, the co-treatment with PD168393 and HRG1β had no impact on the expression of NRG1, as compared to the treatment with HRG1β alone (P>0.05) (Fig. 7C). Moreover, co-treatment with PD168393 and HRG1β also suppressed HRG1β-induced Akt-1, p38MAPK transcription and activation (Fig. 7B, E and F), implying that ErbB2 may serve as downstream factor of NRG1 in the regulation of signaling pathways of both Akt-1 and p38MAPK.