Defeat Osteosarcoma

International Journal of Cancer

Most chemicals were obtained from Sigma (St. Louis, MO) unless otherwise noted. Cell culture media, media ingredients and antibiotics were from Invitrogen (Carlsbad, CA). Bortezomib was from Millennium Pharmaceuticals (Cambridge, MA). Oligonucleotides were custom made by IDT (Coralville, IA). Primary antibodies for Runx2 (mouse monoclonal) and ubiquitin (rabbit polyclonal) were from Santa Cruz (Santa Cruz, CA), for Bax (rabbit monoclonal) from Epitomics (Burlingame, CA), and for β-actin (mouse monoclonal) from Sigma. The Ki-67/active caspase-3 antibody mix for immunohistochemistry was from Biocare (Concord, CA). Secondary antibodies, Precision Plus molecular weight markers, dry milk and buffer ingredients for immunoblotting were from Bio-Rad (Hercules, CA). Secondary antibodies for immunohistochemistry were from Rockland (Gilbertsville, PA). Retroviral shRNA vectors against Runx2 and a control vector, pSM2c, were from Open Biosystems (Huntsville, AL). FuGENE® HD transfection reagent was from Roche (Basel, Switzerland)

Human fetal osteoblasts transformed with SV40 T antigen, hFOB 1.19, and osteosarcoma cells, HOS and 143B, were obtained from ATCC (Manassas, VA). Luciferase-expressing 143B-luc cells were a kind gift of Dr. T.C. He. OS187 cells were a kind gift of Dr. R. Gorlick. HFOB cells were cultured in DMEM/F12 medium at 37°C. HOS, 143B, 143B-luc and OS187 cells were cultured in DMEM medium at 37°C. All media were supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin mixture. Seeding densities were kept constant and adjusted for different plating surface areas. All cells were cultured for 24 hr before treatments. Either bortezomib dissolved in PBS or PBS alone was added to cell media, and various assays were performed at indicated time-points as described below. Subsets of 143B cells were infected with either pSM2c control retrovirus or with 1 of 4 retroviruses carrying different shRNAs against Runx2. Stable clones were selected for 2 weeks using puromycin at 2 μg ml⁻¹. Runx2 expression was assayed with immunoblotting and the 2 clones showing strongest knock-down of Runx2 were chosen for our study along with the stable pSM2c-transfected control.

Cells plated on 6-well plates at indicated time-points were trypsinized, resuspended in 1 ml of PBS and counted using Cellometer® automated cell counter from Nexcelom Bioscience (Lawrence, MA).

Total cellular RNA was isolated using RNeasy Mini Kit by Qiagen (Valencia, CA) and reverse transcribed into cDNA using iScript cDNA synthesis kit by Bio-Rad according to the manufacturer’s instructions. One microgram of cDNA was subjected to real-time PCR using following sets of primers: Runx2 (5-CCG GAA TGC CTC TGC TGT TAT GA-3′ and 5′-ACT GAG GCG GTC AGA GAA CAA ACT-3′), alkaline phosphatase (5′-TGC AGT ACG AGC TGA ACA GGA ACA-3′ and 5′-TCC ACC AAA TGT GAA GAC GTG GGA-3′), Bax (5′- CAC CAG CTC TGA GCA GAT CAT GAA G -3′ and 5′- GCG GCA ATC ATC CTC TGC AG -3′) and GAPDH (5′-GAG TCA ACG GAT TTG GTC GT-3′ and 5′-GAC AAG CTT CCC GTT CTC AG-3′). Real-time PCR was performed using the RotorGene real-time DNA amplification system (Qiagen). SYBR Green reagent produced by Abgene (Rockford, IL) was used to monitor DNA synthesis. The expression of proteins of interest was normalized to the expression of GAPDH.

Cells plated on 12-well plates were transfected with the 6xOSE-luc firefly luciferase reporter9 at 0.5 μg per well. The renilla luciferase promoter-less reporter, pRL-TK, produced by Promega (Madison, WI), was cotransfected at 0.05 μg per well as a reference. FuGene HD reagent by Roche was used for transfections. After 24 hr cells were lyzed and firefly and renilla luciferase activities were measured using a Dual Luciferase Reporter Assay System (Promega) in an Optocomp 1 luminometer according to the manufacturer’s instructions. The firefly luciferase signal was normalized to the renilla luciferase signal and expressed as relative luminescence units (RLU).

Cells were lyzed, and the protein concentration in lysates was measured using the Bradford assay. Twenty five to forty micrograms of total protein per sample was mixed 1:1 with 2xLaemmli buffer; boiled and subjected to electrophoresis using NuPage precast 4–12% gradient polyacrylamide gels (Invitrogen) followed by electroblotting onto polyvinylidene difluoride membranes (Bio-Rad). Blots were blocked in 5% dry milk dissolved in PBS containing 0.01% of Tween-20 (PBST), probed with primary antibody resuspended in 2.5% dry milk dissolved in PBST at 2 (Runx2), 1 (ubiquitin) or 0.5 (Bax) μg ml⁻¹, incubated with horseradish peroxidase-conjugated secondary antibody resuspended in 2.5% dry milk dissolved in PBST at 0.2 μg ml⁻¹, developed using SuperSignal WestPico chemiluminescent substrate produced by Thermo Scientific (Rockford, IL), and photographed. To verify equal loading, blots were stripped in Re-Blot Plus stripping buffer produced by Millipore (Billerica, MA), reprobed with anti-β-actin antibody resuspended in 2.5% dry milk in PBST at 0.5 μg ml⁻¹ and developed as described above. Band intensities were measured using densitometry and Adobe® software.

Cells plated on 24-well plates were treated with bortezomib as indicated in the text. At 24 and 48 hr thereafter, cellular morphology was examined using Zeiss AxioVert inverted microscope under visible light illumination. A nuclear fluorescent stain, Hoechst 33342, was added to the media at 1 μM. Fluorescent nuclear signal was visualized using the above microscope under UV light illumination.

Caspase-3 activity in cell lysates was measured using fluorogenic substrate cleavage assay.25 The reaction mixture contained caspase-3 fluorogenic substrate Ac-DEVD-AMC produced by Calbiochem (San Diego, CA) resuspended at 20 μM in PBS and 10 μg of the cell lysate. The reactions were loaded into 96-well plates and incubated at 37°C for 30 min. Fluorescence of the -AMC tag cut off by active caspase-3 was measured at 440 nm (excitation 380 nm) using a Hitachi plate reader.

To model osteosarcoma in vivo we used human osteosarcoma 143B cells expressing luciferase (143B-luc).26 Cells were grown to confluency and resuspended in sterile Hanks buffered saline (HBSS) at 5 × 10⁶ per ml. Ten microliters of the cell suspension containing 5 × 10⁴ cells were injected orthotopically into the medullar cavity of right tibiae of 5-week-old immunocompromised nude Nu/Nu mice (Crl:Nu/Nu-FoxN1^nu) purchased from Charles River Laboratories (Cambridge, MA). The injection was performed using a Hamilton syringe into the opening in tibial tuberosity pre-made with a 25-gauge needle. Left tibias were sham injected. Tumor growth was monitored longitudinally using bioluminescence imaging (BLI). For bioluminescent imaging the mice were anesthetized and injected intraperitoneally with 100 μl of luciferin substrate. Bioluminescence was measured using a Xenogen imager. A total of 16 mice were injected of which 12 developed tumors as detected with BLI.

On Day 8 after the injection of tumor cells, 12 tumor-bearing mice were randomly divided in 2 groups—the control group (−Bzm) and the treated group (+Bzm). Animals in the treated group received 1 mg kg⁻¹ of bortezomib dissolved in sterile PBS intraperitoneally (IP) every 3 days for 3 weeks. Mice in the control group received sterile PBS IP injections. All animals were sacrificed on Day 28 after the injection of tumor cells, and both hind limbs were excised. Tumor volume was calculated from the 2D caliper measurements using the following formula: tumor volume = length × (width)² × π/6.27

Both tumor-bearing and sham-injected tibias were fixed in 10% neutral buffered formalin for 3 days, decalcified in 14% EDTA for 10 days, and subsequently embedded in paraffin. Serial 4 μm sections were cut and mounted on glass slides. Sections were either stained with H&E or processed for immunohistochemistry as follows: sections were deparaffinized using xylene and ethanol, rehydrated and permeabilized with 0.1% Triton in a TRIS-citrate retrieval buffer at 95°C for 15 min. To block endogenous peroxidase activity, sections were treated with methanol/1% H₂O₂ for 30 min. After the wash, sections were blocked with TBS/0.1% triton/10% normal serum for 1 hr, followed by incubation with the primary antibody for 1 hr. The immunolabeling performed included Ki-67/active caspase-3 double labeling, Runx2 and Bax. After subsequent rinse in PBS, sections were incubated with the corresponding HRP-conjugated secondary antibody, washed and developed with DAB (Ki-67, Runx2 and Bax) and/or Vulcan Red (active caspase-3) substrates. After the development reaction had been terminated, sections were rinsed, dehydrated, mounted and counterstained with either hematoxylin (Ki-67/active caspase-3 and Bax) or FastGreen (Runx2) stains.

To assess the effect of bortezomib on cell proliferation and apoptosis in xenografted osteosarcoma tumors in mice, formalin-fixed, paraffin-embedded tissue sections were probed with the Ki-67/active caspase-3 antibody cocktail. The tissue sections were visualized using a Zeiss Axioplan light microscope equipped with a DAGE video camera that was connected to a Ludl XYZ motorized stage, a color monitor, and a personal computer, and the number of Ki-67 and active caspase-3 positive cells was counted. Quantitative data were obtained utilizing a software program Stereologer 1.3.1 developed by Systems Planning and Analysis (Alexandria, VA), which provided us with an unbiased optical dissector approach for accurate cell counting. The entire tissue section was chosen as a reference area. In addition, spacing of 1,100 μm for Ki-67 and 400 μm for active caspase-3 was established as an optimal interval that would allow us to sample the population of positively stained cells within the reference area. Placement of counting grids provided us with random areas to be analyzed. Results were obtained as the number of positively stained cells per field of view. The treated group was compared to the control group and the results were expressed as the fold change over control (active caspase-3) or percent of control (Ki-67).

Tissue sections probed for Runx2 or Bax were visualized using Zeiss Axioscop 40 microscope and 40× objective. The random fields (10 per section) were photographed under constant illumination and exposure using a high resolution digital camera attached to the microscope. The images were blindly analyzed for nuclear Runx2 or cytosolic Bax staining intensities using the ImageJ software. The treated group was compared to the control group and the results were expressed as the fold change over control.

Experiments were repeated 3 to 5 times. Mean values and standard deviations were calculated, and the statistical significance was determined using a Student’s t-test or ANOVA. Data with p < 0.05 were considered statistically significant.

The majority of osteosarcomas are aggressive and metastatic.6 Human osteosarcoma cell lines, 143B and OS187, have been reported to have an aggressive and invasive phenotype and form tumors and metastases when injected into mice.15, 26, 28 We therefore chose 143B and OS187 cell lines for our study as representatives of an aggressive osteosarcoma. The 143B cells are Ki-Ras-transformed derivatives of HOS cells.26 For this reason, HOS cell line was used as a nontransformed control for 143B cells. HOS cells show limited tumorigenic and metastatic potential.26 OS187 cells were included to expand our study and to exclude a possibility of cell-line specific effects in related HOS and 143B cells. OS187 cell line was produced directly from an osteosarcoma patient specimen.29 It has been described as an aggressive cell line that forms tumors when injected orthotopically into mouse bone and is capable of lung metastasis.28 Immortalized human osteoblastic cell line, hFOB,30 was included as a nonmalignant control. HFOB cells are transformed with a temperature-sensitive SV40 T large antigen which is active at 33°C leading to unrestricted cell growth, but inactivated at 37°C.30 To avoid potential artifacts caused by active SV40 T large antigen, we used hFOB cells at a nonpermissive temperature of 37°C. These conditions allow adequate growth of hFOB cells and have been used in previous studies.31

To assess the phenotype of HOS, 143B and OS187 cells, we first measured their cell growth rate in comparison to hFOB cells. As shown in Figure 1a, the growth rate of HOS cells was slightly higher than the growth rate of nonmalignant hFOB osteoblasts while the growth rate of 143B and OS187 cells was significantly increased. The 143B and OS187 cells had similar doubling time of ∼24 hr. Next, we measured expression of osteoblastic differentiation markers, Runx2 and alkaline phosphatase (ALP) using real-time RT-PCR approach. Figure 1b (top panel) shows that all studied osteosarcoma cells expressed Runx2 confirming their commitment to osteoblastic lineage. Runx2 mRNA levels in HOS and 143B cells were higher whereas in OS187 cells they were 50% lower than in nonmalignant controls. ALP mRNA level (Fig. 1b, bottom panel) in HOS cells was similar to that in hFOB cells while in 143B and OS187 cells, it was significantly decreased, confirming their undifferentiated phenotype. Western blot analysis demonstrated that total Runx2 protein levels were slightly decreased in HOS cells and significantly downregulated in 143B and OS187 cells when compared to hFOB cells (Fig. 1c, top blot). The most pronounced decrease was observed in the faster migrating MASN isoform of Runx2 (lower band) while the slower migrating MRIPV isoform (upper band) was less affected.32 We then assessed Runx2 transcriptional activity using the 6xOSE-luc promoter-reporter construct.9 The basal activity of this reporter in the absence of forced expression of Runx2 was relatively low and ranged from 1,000 to 7,000 RLU or 5- to 20-fold over the background shown as a dotted line in Figure 1d. Figure 1d shows that 6xOSE-luc activity was significantly decreased in all studied osteosarcoma cells when compared to nonmalignant hFOB cells. In our previous work, we have described a regulatory role of Runx2 in Bax expression.18 We therefore investigated whether Runx2 protein levels in the studied cells correlate with Bax protein levels. Figure 1c (middle blot) shows that Bax protein levels correlated with Runx2 levels and were slightly decreased in HOS cells and significantly decreased in 143B and OS187 cells when compared to hFOB cells. In summary, these data indicate that tumorigenic and metastatic osteosarcoma cells, 143B and OS187, have a highly proliferative and undifferentiated phenotype whereas less aggressive HOS cells are less proliferative and more differentiated. Runx2 protein levels correlate with Bax levels and inversely correlate with proliferative capacity in the studied cells.

Our data in Figure 1 illustrate a discrepancy between mRNA and protein levels of Runx2 in the studied osteosarcoma cells. Relative to hFOB cells, in HOS cells Runx2 mRNA was at 260% while Runx2 protein measured with densitometry was at 80%; in 143B cells Runx2 mRNA was at 150% while Runx2 protein was at 46%; and in OS187 cells Runx2 mRNA was at 50% while Runx2 protein was at 38%. This discrepancy between mRNA and protein levels likely indicates increased Runx2 protein degradation. Runx2 is known to be regulated by proteasome20–22; and proteasome inhibition increases Runx2 levels and function.23, 24 We therefore studied effects of a proteasome inhibitor, bortezomib, on Runx2 protein level and function in HOS, 143B and OS187 osteosarcoma cells. For our experiments we chose a concentration of bortezomib of 25 nM because, according to the pharmacokinetic study by Attar et al.,33 this is a physiologically relevant intermediate plasma concentration of bortezomib in patients after a bolus injection. Figure 2a (top blot) shows that 24-hr incubation with bortezomib at 25 nM increased the amount of ubiquitinated proteins confirming the inhibitory effect of bortezomib on proteasomes in the studied cells. Runx2 protein levels were also significantly increased in bortezomib-treated cells (middle blot). Runx2 transcriptional activity as measured with the 6xOSE-luc reporter, was upregulated in bortezomib-treated cells with the most pronounced effect observed in OS187 cells (Fig. 2b). As has been reported by our group before,18 Runx2 regulates Bax expression. Therefore, we examined whether the bortezomib-mediated increase in Runx2 protein level and function correlates with Bax expression. Bax expression in bortezomib-treated osteosarcoma cells was measured using real-time RT-PCR and showed, on average, a 3-fold induction when compared to cells treated with vehicle control (Fig. 2c). The above experiments demonstrate that proteasome inhibition with bortezomib increases Runx2 protein level and activity as well as Bax expression in the studied osteosarcoma cells.

As mentioned above, Runx2 has been shown to function as a tumor suppressor in osteosarcoma. Therefore, by restoring Runx2 protein level and activity, bortezomib may exert an inhibitory effect on osteosarcoma cells. In addition, as a general proteasome inhibitor, bortezomib may affect other molecular pathways regulated by proteasome and thus confer an additional suppression of cell proliferation and functioning. To determine the effect of bortezomib in the studied cells, we performed a dose-response and time-course analysis by evaluating growth rates of nonmalignant hFOB cells and of HOS, 143B and OS187 osteosarcoma cells in vitro. Figure 3a shows that increasing doses of bortezomib did not have any inhibitory effect on the growth of hFOB cells. However, in HOS, 143B and OS187 osteosarcoma cell lines bortezomib suppressed cell growth in a dose-dependent manner. To expand our study and confirm the effect of proteasome inhibition, we used another proteasome inhibitor, MG132, and found that after 24 hr it decreased osteosarcoma cell number on average by 70% and hFOB cell number by 30% (Data not shown). Because MG132 is not an FDA approved compound and because it shows higher toxicity in nonmalignant hFOB cells, we continued our studies using only bortezomib. Bortezomib-treated osteosarcoma cells exhibited morphological features of apoptosis, such as shrinkage and rounding of cell bodies along with condensation of chromatin (Fig. 3b). A specific apoptotic marker protease, caspase-3, was activated in bortezomib-treated osteosarcoma cells (Fig. 3c). These data indicate that proteasome inhibition with bortezomib selectively suppresses growth and induces apoptosis in osteosarcoma cells but not in nonmalignant control cells in culture.

By inhibiting protein degradation, bortezomib can have a wide range of effects on various cell signaling systems other than Runx2. To evaluate a specific contribution of Runx2 in the inhibitory effect of bortezomib on osteosarcoma cells, we performed a stable knock-down of Runx2 in 143B cells using 2 different retroviral shRNA constructs. We then treated these cells as well as control cells transfected with the empty pSM2c retroviral vector with bortezomib at 25 nM for 24 hr. Figure 4a shows that cells transfected with either shRNA1 or shRNA2 against Runx2 contained ∼30% of the amount of Runx2 found in cells transfected with the control vector (pSM). Moreover, while treatment with bortezomib significantly increased Runx2 protein levels in control transfectants (Fig. 4a, lane 2), it did not increase Runx2 protein levels in shRNA-transfected cells (Fig. 4a, lane 4 and 6). Bax expression, as measured with real-time RT-PCR correlated with Runx2 protein levels and significantly increased in bortezomib-treated control transfectants but remained unchanged in bortezomib-treated shRNA1- and 2-transfected cells. Figure 4c shows that the knock-down of Runx2 abolished the inhibitory effect of bortezomib in 143B osteosarcoma cells. These results demonstrate that the inhibitory effect of bortezomib in osteosarcoma cells is at least partially dependent on Runx2.

To study the effect of bortezomib on osteosarcoma development in vivo, we performed xenografting of osteosarcoma cells into nude Nu/Nu mice. Luciferase-expressing 143B (143B-luc) cells26 were injected orthotopically into right tibiae of mice. Development of xenografted tumors was followed longitudinally using bioluminescent imaging (BLI). Figure 5a demonstrates the presence of a strong bioluminescent signal in tumor-bearing mice (top right panel) and absence of a signal in uninjected animals (top left panel). By Day 8, tumor-injected tibiae in 12 out of 16 animals developed consistent BLI signal which increased on average 4-fold when compared to the signal at Day 1, indicating successful engraftment of the tumors. Tumor engraftment was later confirmed histologically (Fig. 5a, lower panels). At Day 8, we divided animals in 2 groups so that each group had equal average BLI signal, and started intraperitoneal PBS (control group) or bortezomib (treatment group) injections at a dose of 1 mg kg⁻¹ every 3 days. In the absence of detailed pharmacokinetic studies of bortezomib in mice, it is difficult to correlate the in vivo and in vitro doses. Previously bortezomib was used at 0.5–2 mg kg⁻¹ in mouse models and gave consistent responses without any noticeable toxicity.34, 35 We, therefore, chose an intermediate dose of 1 mg ml⁻¹ for our animal studies. We have not noticed any toxic effects, i.e., weight loss, in our treated animals. Animals were sacrificed at Day 28, and tumor sizes were determined. Figure 5b shows that bortezomib had a pronounced inhibitory effect on growth of osteosarcoma xenografts in mice. The average size of osteosarcoma tumors in the bortezomib-treated group was only 30% of that in the control group.

To elucidate a possible mechanisms of bortezomib-induced inhibition of osteosarcoma tumor growth, we performed histochemical studies and immunostaining of formalin-fixed paraffin-embedded tumor sections. Cytological analysis of H&E stained sections revealed presence of cells exhibiting apoptotic morphology in bortezomib-treated group, such as shrinkage and rounding of cell bodies along with condensation and fragmentation of nuclei (Fig. 5c, left panel, marked by white arrows). To assess changes in tumor cell proliferation and confirm induction of apoptosis in the bortezomib-treated group in vivo, sections were subjected to double immunostaining for Ki-67 and the active form of caspase-3 respectively (Fig. 5c, right panel). The assay showed decreased number of cells positive for Ki-67 (brown staining marked by yellow arrow) and increased presence of cells positive for active caspase-3 (pink staining marked by black arrows). To determine whether these effects of bortezomib correlated with the increase in Runx2 and Bax levels in xenografted tumors, tumor sections were probed with Runx2- and Bax-specific antibodies (Fig. 5d). The assay showed that both Runx2 (top panels) and Bax (bottom panels) staining intensities were significantly increased in tumors from the bortezomib-treated group. Quantitative analysis of immunostaining for Ki-67, active caspase-3, Runx2 and Bax demonstrated that Ki-67 immunoreactivity decreased by 60% (Fig. 6a); active caspase-3 immunoreactivity increased 3-fold (Fig. 6b); Runx2 immunoreactivity increased 2.7-fold (Fig. 6c) and Bax immunoreactivity increased 2-fold (Fig. 6d) in the bortezomib-treated animals when compared to the control animals. Taken together, these experiments indicate that proteasome inhibition with bortezomib suppresses growth, induces apoptosis, and increases Runx2 and Bax levels in xenografted osteosarcoma tumors in mice.