The human chondrosarcoma cell line SW1353 was purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in Dulbecco’s modified Eagle’s medium/nutrient mixture F-12 (DMEM/F12) with 10% fetal bovine serum (FBS), penicillin (100 U/ml) and streptomycin (100 U/ml) at 37°C with 5% CO₂. HPI-4, also known as Ciliobrevin A, whose molecular formula is C₁₇H₉Cl₂N₃O₂, was purchased from Sigma-Aldrich (St. Louis, MO, USA). The study was approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology (Wuhan, Hubei, China).

To evaluate the change of cell proliferation ability, we used the 3-(4,5-dimethylthiazol-2-yl)- 5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS-8) assay to measure cell proliferation and the toxicity of this drug. The procedure was as follows: 2,000 cells were plated in 96-well plates per well. HPI-4 was added to cells at concentrations of 0, 5 and 10 μM in 100 μl DMEM/F12 with 10% FBS and incubated for 0, 1, 3, 6 and 9 days. Then, 10 μl MST-8 (Beyotime, China) was added to the media in each well and incubated in an environment without light for 90 min. The absorbance value was measured using an enzyme microplate reader at 450 nm wavelength. The relative viability of cells was expressed by OD value.

SW1353 cells were harvested by trypsin enzyme digesting and 200 cells were seeded in a 6-well plate. After cells attached, HPI-4 was added to cells at different concentrations of 0, 5 and 10 μM in each well. Cells were cultured in DMEM/F12 medium with 10% FBS and incubated at 37°C with 5% CO₂ and fresh medium was changed every 3 days. Four weeks later, cells were fixed with 4% paraformaldehyde and stained with 0.2% crystal violet. Finally, images were captured to record the size and number of colonies.

The invasion assay was conducted using Transwell plates. The filter membrane (pore size, 8 μm) of the Transwell plates was coated with 20 mg Matrigel (BD Biosciences, USA) at 37°C for 30 min and the lower chamber was filled with culture medium DMEM/F12 containing 10% FBS. SW1353 cells were starved with serum-free DMEM/F12 for 6 h and then 1×10⁵ cells were transferred onto the upper surface of the chamber. HPI-4 was added at different concentrations of 0, 5 and 10 μM. Twenty-four hours later, the culture medium was gently removed and the upper surface Matrigel was wiped by cotton swab. We used 4% paraformaldehyde to fix the cells that had invaded through and adhered to the lower surface for 15 min, then stained with 0.2% crystal violet. Ten random fields in each plate were selected and counted using an Olympus microscope.

Tumor migratory behavior was imitated with a scratch assay by measuring migration distance. First, SW1353 cells were seeded in 6-well plates at a density of 1×10⁵/well. When they reached nearly confluent monolayer, a single scratch was made vertically through each well by use of a sterile pipette tip. After washing with PBS three times, the cells were incubated with 2% bovine serum DMEM/F12 and cells were stimulated with different concentrations of HPI-4, and observed under a microscope (10× objective lens) after 0, 24 and 48 h. Images were recorded at the same position in order to assess the repair process accurately. Finally, the percentage of cell migration distance was calculated.

We collected 10 human chondrosarcoma tissues from surgical resection and these tissues were firstly fixed in 4% paraformaldehyde. They were then embedded in paraffin and sectioned for immunohistochemical assays. All experimental processes were conducted using standard techniques (12). Ihh (1:100; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), GLI1 (1:400; Epitomics Inc., Burlingame, CA, USA), GLI2 (1:100; Boster, Wuhan, China) and PTHrP (1:100; Santa Cruz Biotechnology, Inc.) were used as markers for activation of HH signaling pathway. Ki67 staining was used to detect proliferation cells in tissue sections at a dilution of 1:200 (Cell Signaling Technology, USA). All sections were observed under a microscope at ×200 magnification.

SW1353 cells were cultured on the cover glass and treated with different concentrations of HPI-4. Twenty-four hours later, they were fixed with 4% paraformaldehyde for 15 min, blocked with 5% BSA at room temperature for 60 min and then incubated with acetylated α-tubulin antibody (1:400; Abcam, Cambridge, UK) overnight at 4°C. Subsequently, the SW1353 cells were incubated with Cy3-conjugated goat anti-Mouse IgG secondary antibody (Boster) at room temperature and nuclei were stained with 1 μg/μl DAPI. We used PBS to wash cells three times during each step during this process. In the end, images were visualized with fluorescent microscope at ×200 magnification.

HH signaling pathway-related gene (GLI-1, PTCH1, Smo, IFT88, PTHrP) expression was measured by RT-PCR. Total RNA was extracted from cells with TRIzol reagent after 48 h incubation with HPI-4 in 6-well plates and then 2–5 μg of total RNA was used to synthesize cDNA with the SuperScript II cDNA synthesis kit (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions. The total PCR system contained cDNA, SYBR-Green, no RNA enzyme water and primers; the primer sequences are listed in Table I.

These procedures were based on previously described ones (6). Total cell proteins (45 μg/lane) were loaded and separated by sodium dodecyl sulfate (SDS) polyacrylamide gels, transferred onto PVDF membranes. We used the following primary antibodies: MMP2 (1:1,000), MMP9 (1:1,000) (both from Cell Signaling Technology), IFT88 (1:1,000; Abgent, USA), CyclinD1 (1:100; Boster) and PTHrP (1:100; Santa Cruz Biotechnology, Inc.); β-actin (1:400; Boster), along with secondary antibody horseradish peroxidase-labeled goat anti-rabbit and goat anti-mouse IgG (1:5,000; Boster). We used ECL western blotting detection kit (Thermo Fisher Scientific, USA) to detect all the protein bands and were visualized using an enhanced chemiluminescence system (Bio-Rad, Hercules, CA, USA). All values were expressed relative to β-actin.

Each experiment was performed at least three times independently. These data are represented as mean ± SD. We used the Student’s t-test or one-way analysis of variance (ANOVA) to analyze the differences between means, and P<0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using SPSS 20.0.

Previous studies demonstrated that a variety of tumors exist with abnormal activation of the HH signaling pathway (7–11). Using immunohistochemistry assay, we detected the expression of the Ihh-PTHrP-related proteins in human chondrosarcoma tissues. These tumor tissues exhibited significantly high level of staining intensity for Ihh, GLI1, GLI2 and PTHrP (Fig. 1B–E) compared with the negative control (Fig. 1A). At the same time, tumor tissues had a considerable percentage of positive Ki67 staining in approximately half of total tumor cells (44.30±4.52%). Obvious expression of Ki67 confirmed that chondrosarcoma had considerable proliferation ability (Fig. 1F). On the other hand, apparent expression of Ihh and PTHrP simultaneously indicated that there appeared no natural negative feedback regulation between Ihh and PTHrP proteins. These results demonstrated that the distinct activation of HH signaling pathway may have a positively correlated relationship with elevated proliferation ability in chondrosarcoma.

As shown in Fig. 2, in order to examine the effect of HPI-4 treatment on human chondrosarcoma cell growth, SW1353 cells were stimulated respectively with 5 and 10 μM HPI-4 and the changes were explored using MTS-8 assay. With the progress of time, SW1353 cells cultured with HPI-4 revealed an obvious reduction in proliferation rate compared with the empty control group (Fig. 2A). In addition, in order to detect the effect of this drug on a single cell to form a tumor mass, we conducted a tablet colony formation assay (Fig. 2B). After 4 weeks of incubation, the empty control group number was 272±19.8 per well compared with 5 μM (191.00±8.49, P<0.05) and 10 μM (90.50±16.26, P<0.05). The average size showed the same trends at different doses (Fig. 2B). These results confirmed that HPI-4 clearly reduced the ability of a single cancer cell to generate a colony by continuous division, including numbers and sizes of colonies (Fig. 3C). Moreover, western blot assay presented a similar conclusion that inhibiting HH signaling eventually suppressed cell cycle protein CyclinD1 in HPI-4-treated cells along with the higher concentration (Fig. 7B). CyclinD1 is considered a vital symbol of cell cycle progression.

Migration and invasion are important indicators to evaluate the degree of tumor malignancy, they determine the severity of the tumor and the prognosis of treatment (7,8). In order to illustrate the influence of HPI-4 compound on tumor migration and invasion ability, we performed the Transwell and cell scratch experiment in vitro. Fig. 3A–D shows that the number of cells penetrating through the Matrigel matrix decreased significantly as the drug concentration increased (the white arrows). Matrigel matrix is regarded as the best component to imitate extracellular matrix. Scratch experiment was used to simulate the migration of tumor cells. The results showed that a higher concentration of HPI-4 observably inhibited tumor cell migration rate as time passed. HPI-4 effectively delayed scratch healing time (Fig. 4A and B). At the same time, we conducted western blot experiment to test the expression of matrix metalloproteinases (MMPs) which reflects tumor cell decomposition of the extracellular matrix ability to promote invasion and migration (22,23). The results demonstrated that after stimulating for 3 days, the expression level of MMP2, MMP9 decreased in HPI-4-treated cells. Moreover, as MMPs upstream signaling pathway, the MAPK/ERK pathway was also restrained by HPI-4, mainly through inhibiting ERK and phosphorylated-ERK (p-ERK) proteins to downregulate downstream MMPs (Fig. 6A–D).

We previously demonstrated that an apparent activation of HH signaling pathway exists, which may be associated with tumorigenesis in human chondrosarcoma (Fig. 1). The mechanism of HPI-4 as an inhibitor of HH signaling pathway is not fully clear, although some studies have demonstrated that this drug does not act on conventional Smo receptor (14). However, primary cilia are an indispensable part of the HH signaling pathway. This cell organelle can be counted and described under the field of vision and was regarded as an observed morphology symbol of cell cycle (15,16). In order to detect whether HPI-4 realizes its function through interacting with primary cilia, we used fluorescence microscope to observe the expression percentage of primary cilia. Normal chondrosarcoma SW1353 cells present primary cilia 20.55±4.40%, while adding 5 μM HPI-4 resulted in a decrease to 15.06±5.15% (P<0.05) and 10 μM to 7.38±1.91% (P<0.001). The results confirmed that HPI-4 distinctly downregulates ciliogenesis, particularly at high concentrations (Fig. 5A–D). Moreover, RT-PCR analysis showed that the expression of cilia microtubule transporters IFT88 significantly decreased (P<0.05) and the HH pathway target genes PTCH1 and GLI1 decreased distinctly (P<0.05), but only the higher concentration of HPI-4 suppressed Smo receptor expression. Thus, we speculated that HPI-4 may inhibit the action of the HH pathway by disrupting ciliogenesis rather than by suppressing Smo (Fig. 7E and F).

HH and PTHrP play an important role in regulating the growth process of bone and cartilage. In normal cartilage tissue, an Ihh/PTHrP negative feedback loop exists. HH signaling pathways activate downstream GLI transcription factors to regulate the generation of PTHrP (2–6). We previously showed that GLI2 and PTHrP have a certain collaborative expression in the human chondrosarcoma tissues with apparent activation of the HH signaling pathway (Fig. 1E). As shown in Fig. 7, after stimulating chondrosarcoma cells with different concentrations of HPI-4, ciliogenesis-related microtubules transporter IFT88 was inhibited, which led to primary cilia dependency. HH pathway transduction was blocked and downregulated transcription factor protein GLI2. Thus, chained along with suppression of PTHrP protein and mRNA expression simultaneously (Fig. 7A–D).