Oncostatin M promotes STAT3 activation, VEGF production, and invasion in osteosarcoma cell lines

Canine OSA cell lines, OSA 8 and 16 were provided by Dr. Jaime Modiano (University of Minnesota, Minneapolis, MN) [6]. The canine D17 OSA cell line and human OSA cell lines U2OS and SJSA were purchased from American Type Cell Culture Collection (ATCC). Cell line authentication of human OSA cell lines SJSA and U2OS was recently completed by The Ohio State University Comprehensive Cancer Center Molecular Cytogenetics Shared Resource through karyotype analysis and comparison to that of the cell lines at ATCC. The canine lines and human line SJSA were maintained in RPMI-1640 supplemented with 10% fetal bovine serum, non-essential amino acids, sodium pyruvate, penicillin, streptomycin, L-glutamine, and HEPES (4-(2-hydroxethyl)-1-piperazineethanesulfonic acid) at 35°C, supplemented with 5% CO2. The U2OS cell line was cultured in McCoy's medium with 10% FBS and the same supplements as listed for the canine lines. The normal canine osteoblasts were obtained from Cell Applications (San Diego, CA) and maintained in Canine Osteoblast Growth Medium with 10% FBS. Human spleen total RNA was purchased from Ambion Biosystems (Austin, TX). The canine OSA tumor and normal spleen samples were obtained from dogs treated at The Ohio State University College of Veterinary Medical Center in compliance with established hospital policies regarding sample collection as part of the Biospecimen Repository. Collection procedures by the Biospecimen Repository are approved by the OSU IACUC (2010A0015, Canine Specimen Collection and Banking). Fresh tissue samples were immediately flash frozen in liquid nitrogen and stored in The Ohio State University College of Veterinary Medicine Biospecimen Repository. The tumors were evaluated and confirmed as OSA by board certified veterinary pathologists at The Ohio State University College of Veterinary Medicine.

RNA was extracted from untreated canine and human OSA cells and pulverized fresh frozen canine OSA tumor samples using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. To generate cDNA, 2 μg of total RNA and the M-MLV reverse transcriptase kit (Invitrogen, Carlsbad, CA) were used according to the manufacturer's instructions. Next, 1/20 of the resultant cDNA was used for each PCR reaction in a total volume of 25 μl. Primers were designed and utilized for canine and human interleukin-6, interleukin-6 receptor, oncostatin M, oncostatin M receptor, gp130, and GAPDH (Table 1). Annealing temperatures for these reactions are listed in Table 1. All PCR products were run on a 2% agarose gel with ethidium bromide and visualized using the Alpha Imager system (Alpha Innotech Corp, San Leandro, CA).

Protein lysates were prepared and quantified, separated by SDS-PAGE, and Western blotting was performed as described previously [6] on 2 × 10⁶ OSA cells after stimulation with either PBS or recombinant human oncostatin M (rhOSM, 50 or 100 ng/mL; R&D Systems, Minneapolis, MN) or recombinant canine interleukin-6 (rcIL-6) (30 ng/mL; R&D Systems, Minneapolis, MN) for 0, 5, 10, or 30 minutes. Additionally, human OSA cell line SJSA was stimulated with either PBS, 50 or 100 ng/mL rhOSM, or 100 ng/mL rhOSM and the small molecule STAT3 inhibitor LLL3 [6] at 40 μM for 72 hours prior to collecting cells and preparing protein lysates that were separated by SDS-PAGE. The membranes were then incubated overnight with anti-p-STAT3 (Y705, Cell Signaling Technology, Danvers, MA), anti-p-JAK2 (Y1007/1008, Millipore, Temecula, CA), anti-VEGF (Santa Cruz Biotechnology, Santa Cruz, CA), or anti-p-Src (Y418, Invitrogen, Carlsbad, CA) after which they were incubated with appropriate horseradish peroxidase linked secondary antibodies, washed, and exposed to substrate (SuperSignal West Dura Extended Duration Substrate, Pierce, Rockford, IL). Blots were stripped, washed, and reprobed for β-actin (Santa Cruz Biotechnology, Santa Cruz, CA), total STAT3 (Cell Signaling Technology, Danvers, MA), total JAK2 (Cell Signaling Technology, Danvers, MA) or total Src (Cell Signaling Technology, Danvers, MA). Images shown are representative of all repeats of the experiments. Experiments were repeated twice.

OSA cells (7 × 10⁶) cells were serum starved for two hours then treated with rhOSM (50 ng/mL) for 0 or 15 minutes. Cells were collected and lysate prepared as described previously [6]. The Rabbit TrueBlot™ kit (eBioscience, San Diego, CA) was utilized to immunoprecipitate canine gp130 using anti-gp130 antibody (Santa Cruz Biotechnology, Santa Cruz, CA) according to manufacturer's instructions. Protein was separated by SDS-PAGE and transferred to a PVDF membrane. Western blotting using an anti-Src or anti-STAT3 antibody (Cell Signaling Biotechnology, Danvers, MA) was performed after addition of the appropriate secondary antibody. The membrane was stripped and reprobed for gp130 and β-actin (Santa Cruz Biotechnology, Santa Cruz, CA).

OSA cells (2 × 10³ cells/well) were seeded in 96-well plates overnight and incubated with PBS, 50, or 100 ng/mL rhOSM for 72 hours. Each treatment group was performed in four replicate wells. Prior to collection, media was removed and the plates were frozen at -80°C overnight before processing with the CyQUANT^® Cell Proliferation Assay Kit (Molecular Probes, Eugene, OR) according to manufacturer's instructions and analyzed as described previously [6].

Cells were plated as previously described [6] and treated with PBS, 50, or 100 ng/mL rhOSM or 100 ng/mL rhOSM and the small molecule STAT3 inhibitor LLL3 40 μM [6]. Separate experiments were conducted with cells plated in a similar manner and treated with PBS, rhOSM (50 ng/mL), rhHGF (50 ng/mL), or the two together. Media was collected after 72 hours, processed, and gel zymography performed as described previously [6]. Images were scanned and analyzed using Image J.

Canine (OSA8) and human (SJSA) OSA cells were plated in invasion assay experiments as described previously [31]. Briefly, cells were plated in the upper chamber in serum-free media with rhOSM (50 ng/mL) for all treatment groups. The lower chambers contained media with 10% fetal bovine serum alone (C10), C10 with rhOSM (50 ng/mL), C10 with rhHGF (50 ng/mL), or C10 with both cytokines at 50 ng/mL. Cells were incubated overnight to allow invasion through the Matrigel layer. Inserts were processed and cells counted as previously described [31]. Treatments were run in quadruplicate and cells from ten random fields from each replicate were counted.

125,000 canine (OSA8) or human (SJSA) OSA cells were plated in C10 media in a six well plate and cultured overnight. The media was removed and cells incubated for 24 hours in C1 media with PBS, OSM 50 or 100 ng/mL, or OSM 100 ng/mL + LLL3 40 μM. Media was removed and frozen at -80°C. VEGF expression was determined using the DuoSet ELISA Development System for canine or human VEGF (R&D Systems, Minneapolis, MN) according to manufacturer's instructions.

In the invasion assays, we computed the average cell count per replicate and analyzed the means using a randomized block ANOVA (blocked on plate). Prior to analysis, the means were square root transformed in order to better satisfy the normality and equal variance assumptions of ANOVA. An overall F test of a difference in means across treatment groups was computed and pairwise comparisons of the groups were performed using Holm's method to control type-I error [32]. All experiments were performed two to three times. Statistical analysis of the VEGF ELISA data was performed using the Student's t-test. P values of less than or equal to 0.05 were considered statistically significant.

Expression of IL-6, IL-6 receptor, OSM, OSMR, and gp130 was determined in three canine (OSA8, 16, and D17) and two human (SJSA and U2OS) OSA cell lines by RT-PCR (Figures 1a and 1b, respectively). All cell lines expressed message for gp130 and OSMR; no expression of OSM was detected. IL-6 expression was variable and weak in canine OSA8 and D17 and human SJSA cells and IL-6 receptor was weakly expressed in canine OSA16 and human SJSA and U2OS cells. Given the apparent lack of IL-6/IL-6R expression in the OSA cells, we focused on OSM and its receptor in the fresh frozen OSA tumor samples from canine patients. OSMR expression was noted in all 8 canine tumor samples evaluated as well as the normal canine osteoblasts while OSM expression was detected in all samples although 2 of these were weak; normal canine osteoblasts did not express OSM. (Figure 1c).

OSM is known to activate the OSMR/gp130 heterodimer leading to phosphorylation of the JAK family kinases, in particular JAK2. Canine (OSA8) and human (SJSA) OSA cell lines were serum starved then stimulated with rhOSM (50 ng/mL) for 0, 5, 10, or 30 minutes before collecting cells for Western blotting (Figure 2). Basal levels of phosphorylated JAK2 were very low in both cell lines, however stimulation with OSM led to an immediate, transient increase in phosphorylation in OSA8 and a more sustained, time dependent increase in SJSA. Basal levels of STAT3 and Src phosphorylation were present as described previously in the OSA cell lines [6]; however, phosphorylation of both STAT3 and Src increased substantially within five minutes of OSM treatment. Levels of total protein for STAT3, Src, and JAK2 remained largely unchanged during all time points.

Given the expression of mRNA for IL-6 receptor in canine OSA cell line OSA16, we wanted to determine whether stimulation with its ligand IL-6 would impact JAK2 or STAT3 phosphorylation as had occurred with OSM. Cells were serum starved then treated with rcIL-6 for 0, 5, 10, or 30 minutes before cells were collected for Western blotting. JAK2 phosphorylation was not present at baseline and stimulation with IL-6 did not induce JAK2 phosphorylation (Figure 3). Basal STAT3 phosphorylation was present in OSA16 and this was not altered following IL-6 stimulation. Levels of total STAT3 and JAK2 proteins were not altered during all time points evaluated.

Binding of OSM to its receptor and gp130 results in recruitment of JAK2 to the receptor complex and subsequent recruitment and phosphorylation of STAT3. This association likely explains the activation observed in Figure 2, however the activation of Src after OSM binding is not as clear. Multiple members of the Src family of tyrosine kinases were coprecipitated with gp130 in lysates from multiple myeloma cells and stimulation with IL-6 led to increased activity of these Src family kinases [26]. We wanted to determine whether Src was associated with the gp130 complex in OSA cells as well. Canine (OSA8) and human (SJSA) OSA cell lines were serum starved for 2 hours then left untreated or treated for 15 minutes with rhOSM (50 ng/mL). Lysates were collected and gp130 was immunoprecipitated from the canine and human OSA cell lines (Figure 4). Western blotting revealed that Src and STAT3 were associated with gp130 in the presence or absence of OSM indicating that these proteins are part of the gp130 complex in these cell lines. The lack of β-actin in the co-precipitates confirmed the specificity of the immunoprecipitation experiment (Figure 4).

OSM is a cytokine with multiple, divergent effects on cell proliferation differing among cell types and lines with growth inhibition effects reported in melanoma and glioma cells but stimulation of growth of Kaposi's sarcoma cells [11]. Canine (OSA8) and human (SJSA) OSA cell lines were incubated with 0, 50, or 100 ng/mL rhOSM for 72 hours and proliferation was assessed using the CyQUANT assay. As shown in Figure 5, there was no effect of OSM stimulation on OSA cell proliferation at either concentration.

Previous work has shown that OSM promotes expression of MMPs including MMP1 and MMP3 in astrocytes [33], MMP1 and MMP9 in fibroblasts [33], and MMP1, MMP3, and MMP13 in chondrocytes [34]. Indeed, increased expression of MMP2 and MMP9 was linked to increased invasive capability in human and canine OSA [35, 36]. We treated canine (OSA8) and human (SJSA) OSA cell lines with 0, 50, or 100 ng/mL rhOSM or 100 ng/mL OSM and 40 μM of the small molecule STAT3 inhibitor LLL3. We have shown in previous work that this STAT3 inhibitor down-regulates MMP2 expression at 72 hours following exposure [6]. OSM stimulation induced a dose dependent increase in MMP2 activity that was abrogated in the presence of LLL3 suggesting that the increase in MMP2 activity conferred by OSM stimulation is due in part to STAT3 activation (Figure 6a).

To determine whether the effect of OSM on MMP2 expression was biologically relevant with respect to tumor cell invasion, we cultured canine (OSA8) or human (SJSA) OSA cells in inserts containing serum-free media and rhOSM (50 ng/mL) overlying a Matrigel substrate. These inserts were placed in wells containing either media with 10% fetal bovine serum alone (C10), C10 with rhOSM (50 ng/mL), C10 with rhHGF (50 ng/mL), or C10 media with both cytokines together at same concentrations (50 ng/mL). After 18 hours of incubation, OSA cell lines treated with either cytokine alone exhibited significantly enhanced invasion as compared to media alone. (Figure 6b). Furthermore, invasion of OSA cells treated with both rhOSM and rhHGF was significantly greater than that observed with either cytokine/growth factor alone. Upregulation of MMP2 activity was observed following treatment with rhOSM alone, rhHGF alone and both OSM and HGF in combination (Figure 6c). Finally, stimulation of the human OSA cell line SJSA with OSM led to dose dependent increases in VEGF protein expression that was largely abrogated by concurrent treatment with the small molecule STAT3 inhibitor LLL3 (Figures 6d and 6e).