Experimental data and limited patient experience suggest that rhBMP-2 can be used to regenerate bone in acquired segmental defects of the mandible. Most of these defects are caused by resection of oral squamous cell carcinoma (OSCC) and the biologic effects of rhBMP-2 on these carcinoma cells are unknown. The objective of this study was to determine whether rhBMP-2 produces adverse effects on proliferation and angiogenesis in OSCC, two biologic processes critical to tumor formation. In vitro studies included treating OSCC cells with rhBMP-2 or an adenoviral vector containing the cDNA for BMP-2. In vivo studies involved co-transplantation of OSCC cells with bone marrow stromal cells genetically modified to over express BMP-2, to mimic a clinically relevant scenario for regenerating bone using cell-based therapy in a wound containing microscopic residual disease. Proliferation, as measured by a MTT assay in vitro and tumor growth in vivo was not affected by treatment with BMP-2. Angiogenesis, measured by secretion of the proangiogenic molecules VEGF and IL-8 in vitro and microvessel density in vivo, was not affected. Exposure of OSCC cells to BMP-2 does not stimulate proliferation or angiogenesis. Further studies are needed before using rhBMP-2 for bone tissue engineering in oral cancer-related defects.
In the craniofacial region, large defects that require bone reconstruction most commonly arise from resection of advanced-stage malignancies. These defects are less often caused by excision of benign neoplasms, osteoradionecrosis, osteomyelitis, trauma or congenital disorders. The use of autogenous non-vascularized bone graft is the method of choice to repair small bone defects. The gold standard for large defects is now vascularized free tissue flaps. This approach successfully restores form and function when performed by appropriately trained surgeons, but free tissue transfer has limitations, including donor site morbidity, difficulty restoring the complex 3-dimensional structure of the defect, significantly extending the surgical time, and a 5–10% failure rate due to vascular thrombosis . These limitations have led to newer regenerative approaches using tissue engineering concepts being considered for reconstructing these defects.
Extensive preclinical studies in various animal models have demonstrated successful healing of critical-sized mandible and calvarial defects using osteoinductive bone morphogenetic proteins (BMPs) delivered on biodegradeable carriers or by a gene therapy vectors . Based on the success of multicenter prospective clinical trials, regenerative therapy using recombinant human BMP-2 (rhBMP-2) is now FDA-approved for spinal fusion, alveolar ridge augmentation and sinus floor augmentation . BMPs can be used to heal craniofacial bone defects associated with cleft palate, LeFort advancement surgery and segmental mandibulectomy .
Although BMPs were discovered because of their osteoinductive activity, these proteins have important roles during embryogenesis, such as the development of the teeth, brain, eyes, heart, kidneys, gut, lung, skin, bone and testis . BMPs are constitutively expressed by some malignancies including melanoma, pancreatic, lung, prostate and breast cancers . The role of the BMP-signaling pathway in carcinogenesis remains poorly understood, but there is concern about the possibility of adverse effects such that the use of rhBMP-2 in cancer patients is contraindicated. The biological effects of BMP-2 on oral squamous cell carcinoma (OSCC) must be studied before using this molecule for reconstructing bone defects caused by oral cancer. The purpose of this study was to determine the in vitro and in vivo effects of BMP-2 on proliferation and angiogenesis in OSCC cells; two functional capabilities of cancer cells critical for carcinogenesis .
Materials and methods
The human OSCC cell lines UMSCC-1 (floor of mouth), and UMSCC-74A (tongue) were obtained as a gift from Dr Thomas Carey (University of Michigan). Cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1% penicillin and streptomycin (Invitrogen, Carlsbad, CA, USA).
Human bone marrow stromal cells (BMSCs) were collected from the iliac crest of patients undergoing reconstructive surgery that required harvest of iliac crest non-vascularized bone graft . Cells were maintained in alpha modified Eagle’s medium (α-DMEM) supplemented with 15% fetal bovine serum, 1% penicillin and streptomycin. All procedures were performed with an approved protocol from the University of Michigan Institutional Review Board.
Cell Proliferation Assay
Cells were grown as described and 2500 cells were seeded onto each well of 96-well tissue culture plates. The following day, cells were treated with one of the following: transduction with an adenoviral vector containing the cDNA for BMP-2 (AdCMVBMP-2) at multiplicity of infection (MOI) of 0, 10, 100, 500 or 1000; transduction with a control adenoviral vector containing the cDNA for lac-z (AdCMVlacZ) at MOI of 0, 10, 100, 500 or 1000; treatment with rhBMP-2 protein (R&D Systems, Minneapolis, MN) at 0, 20, 100, 250 or 500 ng/ml. Viral transductions were performed by exposure of the cells to the adenoviral vector suspended in DMEM for 24 h. The number of viable cells was determined at 24, 48, 72 and 96 h post treatment using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega, Madison, WI, USA). The absorbance of the plates was recorded at 490 nm.
Vascular endothelial growth factor (VEGF) and interleukin-8 (IL-8) assays
Supernatant from the treatment groups was collected prior to the cell proliferation assays and used to quantify VEGF and IL-8 levels using Quantine ELISA kits (R&D Systems). The plates were read at 450 nm with a correction at 540 nm.
In vivo xenografts
All procedures involving animals were performed in accordance with protocols approved by the Unit for Laboratory Animal Medicine at the University of Michigan. Four week old female NIHS-bg-nu-xid mice were purchased from Charles River Laboratories (Charles River, Wilmington, MA, USA) and housed under pathogen-free conditions. There were two treatment groups: group 1, 1.5 × 10 6 UMSCC-74A cells co-injected with 1.5 × 10 6 hBMSCs (passage 6); and group 2: 1.5 × 10 6 UMSCC-74A cells co-injected with 1.5 × 10 6 hBMSCs which were transduced with AdCMVBMP-2 at 200 MOI for 24 h prior to injection. The cells were suspended in 200 μl of 2 mg/ml type I rat tail collagen (BD Biosciences, Bedford, MA, USA) for injection. After the mice were anesthetized, cells were injected into the right or left dorsal region and 7 mice were killed at 2, 3 and 4 weeks. The tumors were surgically removed, weighed on an electronic balance and measured using a digital caliper. The tumor volume was calculated using the formula volume = (length × width 2 )/2 . Tissue samples were fixed in buffered zinc formalin (Z-Fix, Anatech Ltd, Battle Creek, MI, USA) overnight, embedded in paraffin and serial sectioned at 4 μm thickness.
Immunolocalization of Von Willebrand factor
Tissue sections were treated with peroxide blocking solution for 5 min after antigen retrieval, incubated with primary antibody (anti-Von Willebrand factor, Dakocytomation, Denmark) for 1 h, incubated for 30 min with horseradish peroxidase (HRP) labeled polymer (Dakocytomation EnVision+ System, Kit, CA), and then incubated with 3-amino-9-ethylcarbazole (AEC)-HCl + substrate-chromogen solution for 3 min at room temperature. The number of stained vessels was counted in five random high-power fields at × 20 magnification in a blinded fashion.
The results of cell proliferation, ELISA, tumor volume, and tumor weight are presented as mean ± standard derivation (SD). Student’s t-test was used for comparison between two groups and ANOVA was used for multiple comparisons. Differences were determined as statistically significant at p < 0.05.
Effects of rhBMP-2 and AdCMVBMP-2 on oral carcinoma cells in vitro
Treatment of UMSCC-1 and UMSCC-74A with rhBMP-2 at concentrations ranging from 0 to 500 ng/ml did not have a significant effect on cancer cell proliferation as measured by the MTT assay ( Fig. 1 a, b ). Similarly, transduction of UMSCC-74A cells with AdCMVBMP-2 or AdCMVLacZ at a MOI from 0 to 1000 did not affect cell proliferation ( Fig. 1 d, f). Cell proliferation increased slightly when UMSCC-1 cells were transduced at 10 and 100 MOI with AdCMVBMP-2 or AdCMVLacZ ( Fig. 1 c, e). Transduction of UMSCC-1 cells with AdCMVBMP-2 at higher MOIs (500 and 1000 MOI) reduced cell proliferation and transduction at these high MOIs for AdLacZ did not affect proliferation ( Fig. 1 c, e).
To determine if BMP-2 treatment enhances the secretion of proangiogenic factors known to support tumorigenesis, cell culture supernatants from the rhBMP-2 treated cancer cells were analyzed for the secretion of IL-8 and VEGF. For UMSCC-1 cells, both IL-8 and VEGF secretion decreased up to 4 days after treatment compared with controls ( Fig. 2 a, c ). IL-8 secretion by untreated UMSCC-74A cells decreased over time. This trend was also observed for UMSCC-74A cells treated with rhBMP-2, although the IL-8 levels at each time point tended to be higher than the control group ( Fig. 2 b). For the UMSCC-74A cells, VEGF secretion increased over time in culture for the control and treatment groups. The secretion of VEGF was not significantly increased by BMP treatment over control groups at any time point ( Fig. 2 d).
Effects of BMP-2 on oral carcinoma cells in vivo
To determine the effects of BMP-2 on tumor growth in vivo , the authors carried out subcutaneous injections of UMSCC-74A cells co-transplanted with BMSCs alone (group 1) or with BMSCs that were transduced with AdCMVBMP-2 (group 2). In group 2, there were small amounts of new bone formation on the tumor edge, indicative of biologically active BMP-2 secreted by the transduced cells. There were no statistically significant differences in wet weight or volume between the tumors of UMSCC-74A cells co-transplanted BMSCs with or without AdCMVBMP-2 infection ( Figs. 3a–c ). Similarly, quantification of vessel density using immunohistochemical staining for Von Willebrand factor showed that there were no significant changes between the tumors with and without AdCMVBMP-2 infection at each time point ( Fig. 3 d).