Secondary Logo

Journal Logo

SECTION I: SYMPOSIUM I: Papers Presented at the 2005 Meeting of the Musculoskeletal Tumor Society

VEGF and BMP Expression in Mouse Osteosarcoma Cells

Weiss, Kurt, R; Cooper, Gregory, M; Jadlowiec, Julie, A; McGough, Richard, L, III; Huard, Johnny

Author Information
Clinical Orthopaedics and Related Research: September 2006 - Volume 450 - Issue - p 111-117
doi: 10.1097/01.blo.0000229333.98781.56


Osteogenic sarcoma (OS) is the most common primary malignancy of bone. It affects mainly youths in their second decade of life, and there is a 1.5:1 male to female ratio. Osteogenic sarcoma is a rare neoplasm and there are only between 500 and 700 new cases in the US per annum.3,5,19,22,26,29,32

Before the advent of chemotherapy, only 10% to 20% of patients experienced 5-year disease-free survival with surgical treatment alone. Most patients with OS died secondary to overwhelming pulmonary metastatic disease. Osteogenic sarcoma displays aggressive metastatic potential and more than 95% of metastatic disease occurs in the lung.3,5,19,22,26,29,32 Current treatment protocols for OS include the use of neoadjuvant chemotherapy, wide surgical resection versus amputation, and postoperative maintenance chemotherapy.3,7,10,23,26 Advances in each of these modalities have led to dramatic prognostic improvements for patients with OS. Today, 60% to 70% of patients with OS experience 5-year disease-free survival.26,29,32 Despite these improvements, approximately one third of OS patients still succumb to overwhelming pulmonary metastatic disease refractory to chemotherapy and surgery. The prognosis for patients with radiographically apparent metastatic disease at the time of diagnosis is particularly poor.26 Due to the failure of current modalities to successfully treat patients with metastatic disease, novel treatment options are required.

The bone morphogenetic proteins (BMPs) were first described in 1965 when Urist reported demineralized bone matrix was osteoinductive.30 Urist et al31 later reported the osteoinductivity of devitalized osteosarcoma, which implied both neoplastic and normal bone respond to similar molecular signals. It has been recently suggested BMP expression in OS correlates positively with the incidence of metastatic disease and negatively with disease-free survival, although the BMP expression of metastatic OS cells has not been characterized.1,21,27,33,34 In addition to BMPs, vascular endothelial growth factor (VEGF), a potent neoangiogenic agent widely investigated in oncology, has been implicated in the occurrence of metastasis,8,11,16,25,28 and high concentrations of VEGF in the sera of patients with OS have been linked to the development of pulmonary metastasis.15

Noggin is an antagonist of BMP types 2, 4, and 7.4,35 Noggin physically binds to BMP molecules and prohibits their interaction with the BMP cell surface receptors.12 Decreased noggin signaling has been implicated in several pathologies, including tarsal coalition and fibrodysplasia ossificans progressiva.4 Previous research performed in our laboratory has demonstrated noggin can completely abrogate BMP-mediated ectopic bone formation.13 However, the effects of noggin treatment on the behavior of OS cells have not been investigated.

We hypothesized increased metastatic potential of mouse osteosarcoma cells correlate with the expression of VEGF and BMPs.


We characterized the expression patterns of VEGF and BMPs in two cell lines derived from a single tumor found in a Balb-C mouse. These cells have been designated K7M2 (highly metastatic) and K12 (less metastatic).17,18 By virtue of their common origin but vastly divergent behaviors and metastatic potential, the K7M2 and K12 cell lines serve as unique tools with which to study OS growth and metastasis (Fig 1). K7M2 and K12 cells were provided by Lee Helman, MD, and Chand Khanna, DVM, PhD (Pediatric Oncology Branch, National Cancer Institute, National Institutes of Health). Cells were cultured with Dulbecco's Modified Eagle Medium (Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin (basal medium). They were maintained at 37°C and 5% CO2 at approximately 50% to 75% confluency. On average, they were passaged every 3 days with a range of 2 to 4 days.

Fig 1A
Fig 1A:
B. Murine K7M2 and K12 cell lines have different metastatic potentials. (A) K7M2 cells implanted into the mouse hind limb grow into a large tumor that leads to extensive pulmonary metastases. (B) The K12 cells also form a large primary tumor, but the tumor generates few or no pulmonary metastases.

The QIAGEN RNeasy kit (QIAGEN Inc, Valencia, CA) was used according to the manufacturer's specifications to extract RNA from K7M2 and K12 cells in culture in triplicate. RNA then was quantified with spectrophotometry. Five micrograms of total RNA was used to make cDNA with the SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA), as detailed in the manufacturer's specifications.

Polymerase chain reaction (PCR) primers for murine BMP2, −4, and −7 and β-actin were obtained by a literature search and confirmed with the NCBI online database.24 Primers were manufactured by Integrated DNA Technologies, Inc (Coralville, IA) as follows: BMP2, forward primer: 5′CCAGGTTAGTGACTC AGAACAC3′, reverse primer: 5′TCATCTTGGTGCAA AGACCTGC3′; BMP4, forward primer: 5′TAGCAAGAGTGC CGTCATTCC3′, reverse primer: 5′CCAGTCTCGTGTCCAGT AGTCG3′; BMP7, forward primer: 5′CGATTTCAGCCTGGA CAACG3′, reverse primer: 5′CCTGGGTACTGAACACGG3′; and β-actin, forward primer: 5′GTGGGCCGCTCTAGGCACCA A3′, reverse primer: 5′CTCTTTGATGTCACGCACGATTTC3′. Polymerase chain reactions were performed using the Super-Script First-Strand Synthesis System (Invitrogen) for RT-PCR on a Perkin Elmer GeneAmp PCR System 9700 thermal cycler (Perkin Elmer, Wellesley, MA). The PCR protocol was as follows: 5 minutes at 94°C; 25 cycles of 30 seconds at 94°, 30 seconds at 52°C, and 1 minute at 72°C; and 10 minutes at 72°C. PCR products were loaded onto a 1% agarose gel containing ethidium bromide, visualized with ultraviolet light, and photographed.

The RNeasy Kit and DNase I were obtained from Qiagen. Quantitative real-time PCR master mix reagents and 18S primers and probe were designed and purchased from Applied Biosystems (Foster City, CA). All target gene primers and probes were purchased from Integrated DNA Technologies, Inc.

Cells were seeded in triplicate in complete basal medium in 35-mm plates at approximately 70% confluence. The RNeasy kit with DNase I digestion were used to extract total RNA after 24 hours. After RNA extraction, qPCR analysis was carried out as described previously.14 Sequences for all target gene primers and probes were as follow: Bmp-2 (NM 007553), forward primer: 5′CCAAAATCCCTAAGGCATGCT3′, reverse primer: 5′TTCATTTTCATCTAGGTACAACATGGA3′ and Taqman® probe: 5′TGTCCCCACAGAGCTCAGCGCAAT3′; Bmp-4 (NM 007554), forward primer: 5′ATGTGAGCCCTGCAGTC CTT3′, reverse primer: 5′CTCAATGGCCAGCCCATAAT3′ and Taqman® probe: 5′CTGGACCCGGGAAAAGCAACCC3′; and Bmp-7 (NM 007557), forward primer: 5′TGGCACGTGAC GGACAAG3′, reverse primer: 5′GGCAGACATTTTTCCTGC TCTT3′ and Taqman® probe: 5′CCTACCAGCTACCACAGCA AACGCC3′. Gene expression levels were calculated based on the ΔCT method (separate tubes).6 All target genes were normalized to the reference housekeeping gene, 18S.

K7M2 and K12 cells were grown in Costar six-well plates (Corning, Inc, Corning, NY). At approximately 75% confluence, the cells were washed with phosphate buffered saline (PBS) and given 3 mL of fresh basal medium. Forty-eight hours later, the medium was collected and cells were counted with a hemocytometer. The culture media were then tested with Quantikine quantitative colorimetric enzyme-linked immunosorbent assays (ELISAs) for murine VEGF and BMP2 according to the manufacturer's specifications (R&D Systems Inc, Minneapolis, MN). Optical absorbance of samples was determined with a microplate reader and expressed in graphical format with Microsoft® Excel software (Microsoft Corp, Redmond, WA). Each sample was assayed in duplicate or triplicate.

K7M2 and K12 cells were plated in six-well plates at a density of 5000 cells/well. Fifty percent of the wells were treated with noggin-conditioned medium. Noggin-conditioned medium was harvested from HT1080 cells transduced with a noggin-expressing vector. A noggin bioassay was used to determine the level of noggin produced by the transduced cells as previously described.13 Conditioned medium was diluted to a final concentration of 167 ng/mL for K7M2 and K12 cell line treatments. A microscopic imaging system was used to obtain time-lapse imaging at 100× magnification. This system, which has been previously described,9 consists of a biobox incubator mounted to the stage of a microscope and digital camera. The biobox (Automated Cell Technologies, Inc, Pittsburgh, PA) maintained sterility and a constant temperature and CO2 concentration on top of the microscope stage. The Nikon Eclipse TE 2000 microscope (Nikon, Inc, Melville, NY) was programmed with the coordinates of fields of view, selected on the basis of cell presence, and rotated through them such that each field was under view at 5-minute intervals. Eight fields per well were chosen on the basis of the presence of cells. We ensured we were not filming blank regions of the plate, hence a random selection, some near the center of the plate, some near the edge. The Photometrics Cool Snap ES digital camera (Roper Scientific, Tucson, AZ) collected JPEG images at each field and time point. The plate was imaged in this fashion for 96 hours. The AVI Constructor software then was used to combine the individual JPEG images and construct movies.2 Each movie comprised 1152 individual JPEG images collected at 5-minute intervals for 96 hours.

After the 96-hour noggin treatment was completed, the six-well plate was recovered from the imaging apparatus. Cells were washed with PBS, trypsinized, and counted on a hemocytometer in the presence of trypan blue, a stain taken up by necrotic but not viable cells.

The result of each experiment is reported as the mean of three treatment triplicates ± standard error of the mean (SEM). For qPCR assays, the coefficient of variation (COV) was calculated from 3 assay replicates. The COV did not exceed 3% for any of the analyzed treatment groups or target genes. One-way analysis of variance (ANOVA) followed by Tukey-Kramer's post hoc test (performed with SYSTAT 9 software [SYSTAT Software, Inc, Richmond, CA]) was used to determine significance among treatment groups. A p value less than 0.05 was considered statistically significant.


We observed higher expression levels of BMP2 mRNA and BMP4 mRNA in the highly metastatic K7M2 cells than in the less metastatic K12 cells (Fig 2A). K7M2 cells displayed a band for BMP2 at the expected size of 181 base pairs, whereas K12 cells did not display a BMP2 band. K7M2 and K12 displayed bands for BMP4 at 402 base pairs, with the band from K7M2 appearing much denser than the band from K12. Neither K7M2 nor K12 cell lines showed bands for BMP7 at the expected size of 253 base pairs. The negative (no DNA) and positive (β-actin) controls were appropriately negative and positive in both cell lines (Fig 2A).

Fig 2
Fig 2:
The highly metastatic K7M2 and less metastatic K12 cell lines show different BMP expression patterns by RT-PCR (A) and qPCR (B). L = ladder; 1) BMP2; 2) BMP4; 3) BMP7; 4) Neg. control; 5) β-actin. Bars represent mean ± SEM, n = 3, *Significant from K7M2, p < 0.05, ND = not detected.

The K7M2 cells express higher (p < 0.05) levels of BMP2 mRNA than do K12 cells. However, BMP4 expression seems to be higher (p < 0.05) in the K12 cells. The qPCR data support the RT-PCR data that show undetectable expression of BMP7 in either K7M2 or K12 cells (Fig 2B).

K7M2 cells produced more BMP2 (887 pg per million cells per day) than did K12 cells (no detectable amount) (Fig 3).

Fig 3
Fig 3:
We used a BMP2 ELISA to compare K7M2 and K12. These data represent the average of three experiments and largely parallel the PCR data.

K7M2 cells produced more VEGF (478 pg per million cells per day) than did K12 cells (191 pg per million cells per day) (Fig 4). This represents a 2.5-fold difference in VEGF protein production between the highly metastatic and less metastatic cell lines.

Fig 4
Fig 4:
We performed a VEGF ELISA to compare K7M2 and K12. These data represent the average of two experiments and illustrate K7M2 cells produce more than 2.5-fold more VEGF per million cells per day than K12 cells.

Overall, the highly metastatic K7M2 cells expressed high levels of BMP2 and VEGF whereas the less metastatic K12 cells produced little or none of these growth factors.

Qualitatively, the K7M2 control group cells (without noggin-conditioned medium) were extremely large and motile. At the end of the 96-hour experiment, the fields of view were approximately 50% confluent (Fig 5A). K7M2 cells treated with noggin-conditioned medium (167 ng/mL) were remarkably less motile and appeared qualitatively smaller than the control group cells. There was a small amount of cell death and cells were approximately 50% confluent at the end of the 96-hour experiment (Fig 5A).

Fig 5A
Fig 5A:
B. (A) The effects of noggin treatment on highly metastatic K7M2 cells and less metastatic K12 cells are shown. Compared with untreated controls, noggin-treated K7M2 cells showed markedly decreased motility, qualitatively smaller overall size and smaller cytoplasmic projections, with approximately 10% cell death. Noggin-treated K12 cells demonstrated approximately 30% cell death and decreased confluence (~40% versus ~95%). (B) Trypan blue staining (stain, trypan blue; original magnification, ×40) demonstrated the effect of noggin treatment on K7M2 and K12 cell viability.

K12 control group cells (without noggin-conditioned medium) were much smaller than the K7M2 control cells with fewer and shorter cytoplasmic projections. They were also less motile than the K7M2 control cells. After 96 hours the fields of view were approximately 95% confluent (Fig 5A). Noggin-treated K12 cells showed widespread cell death throughout the entire 96-hour experiment. At the conclusion of the experiment, approximately 1/3 of the cells appeared to be dead and the fields of view were approximately 40% confluent (Fig 5A). There were no major differences in terms of motility or morphology between the control and noggin-treated K12 cells. Quantitatively, control K7M2 cells showed no cell death, whereas noggin-treated K7M2 cells displayed 11.54% cell death (Fig 5B). Control K12 cells had 3.85% cell death, but noggin-treated K12 cells showed 28.57% death (Fig 5B).


Approximately 1/3 of patients diagnosed with OS eventually succumb to pulmonary metastatic disease refractory to surgery and chemotherapy. The tenacity with which OS metastasizes to the lung presents the greatest obstacle to disease-free survival in these patients. Treatments that inhibit metastasis-associated factors in OS cells and thus render them less metastatic would potentially extend the lives of these patients. The role of VEGF and BMP in the metastatic potential of OS is poorly understood. This study was designed to better characterize the growth factor expression by OS cells and to test the effectiveness of a specific therapy.

The major limitation of our study is that it was performed completely in vitro; therefore, the correlation between growth factor expression and metastatic potential must rest on the previous studies involving the characterization of these cells' metastatic potential. Fortunately, these cells have been well characterized in previous studies,17,18 so we believe our conclusions are sound. Furthermore, because our primary goal was to test the effect of noggin therapy on OS cells, in vivo studies are beyond the scope of this study.

K7M2 and K12 are related cell lines derived from the same original tumor. K7M2 is highly metastatic to the lung in a previously described animal model, whereas K12 is much less metastatic.17,18 These phenotypic differences make K7M2 and K12 powerful tools with which to investigate factors that may confer metastatic potential. Here we describe differences in VEGF and BMP expression between highly metastatic (K7M2) and less metastatic (K12) cell lines. We also describe changes in the motility, morphology, and viability of these cells when treated with the BMP antagonist noggin.

The importance of VEGF in tumors has been well described.8,25,28 Theoretically, all tumors must express neo-angiogenic factors if they are to sustain themselves, grow, invade local structures, and metastasize. The expression of VEGF and other neoangiogenic factors have been shown to be upregulated in angiogenic chondrosarcoma.20 Kaya et al found a positive correlation between patients' serum VEGF levels and their incidence of pulmonary metastases in OS, suggesting VEGF is a metastasis-associated factor in OS.15 Our ELISA results indicate the highly metastatic K7M2 cells produce well over twice the amount of VEGF produced by metastatic K12 cells. This data supports the hypothesis that VEGF production correlates with metastatic potential and suggests VEGF antagonists may represent an effective method of therapeutic intervention.

Urist first suggested the importance of BMPs in osteosarcoma more than a generation ago,31 but researchers have only recently reported some OS primary tumors are BMP-positive and BMP expression correlates with the incidence of pulmonary metastasis.1,33,34 We report here K7M2 and K12 have different BMP expression patterns. Our data show robust BMP2 mRNA and protein expression in the highly metastatic K7M2 cells but little or no BMP2 mRNA or protein expression in the less metastatic K12 cells. The strong validation of our BMP2 data by PCR and ELISA is encouraging and gives us confidence that BMP2 is likely a metastasis-associated factor in OS. Conversely, metastatic potential does not seem to correlate with high levels of BMP4 or BMP7 expression. Like the inhibition of VEGF, the inhibition of BMP2 in osteosarcoma may represent a potential avenue of specific antimetastatic therapy.

We examined BMP types 2, 4, and 7 because they are the primary BMPs responsible for osteoinduction and are types naturally inhibited by noggin. Noggin binds to BMP molecules at the receptor binding site, thus prohibiting interaction between BMPs and their receptors.12 Having determined K7M2 and K12 cells have different BMP expression patterns, we treated these cells with noggin and demonstrated noggin had different effects on the two cell lines over time. Treatment with noggin seemed to decrease the motility and size of the K7M2 cells and elicit a small but appreciable decrease in viability. The noggin-treated K12 cells, in contrast, displayed abundant cell death and were not nearly as confluent as the control K12 cells at the 96-hour time point. This response suggests that in addition to being growth-and metastasis-associated factors in OS, BMPs may be survival factors as well, the inhibition of which causes demise of the cells.

Perhaps the reason noggin treatment had a more dramatic effect on K12 cells than on K7M2 cells has to do with their differing BMP expression patterns. According to our findings, K7M2 cells produce more BMP2. Therefore, a high concentration of noggin may be required to alter BMP signaling and/or kill these cells. The K12 cells, in contrast, produce less BMP2, the activity of which may be more easily counteracted by a lower concentration of noggin. This series of experiments provides support for our hypothesis that the inhibition of metastasis-associated factors will fundamentally alter the biology of OS cells and decrease the incidence of pulmonary metastases. This strategy should be investigated further as a possible means to eradicate OS pulmonary metastases.

The results of this study suggest increased BMP2 and VEGF expression may be related to the metastatic potential of OS cells. The findings also demonstrate anti-BMP therapy (with noggin) may be useful for the inhibition of OS metastasis. Future directions could include studies investigating BMP inhibition with variable noggin concentrations, the effects of VEGF inhibition through the use of sFlt to block VEGF signaling,11 or a combination of both. Translation of these in vitro results into an animal model will provide a necessary first step to investigating the in vivo effects of these interactions and to establishing a therapeutic animal model designed to decrease the incidence of OS pulmonary metastases (Fig 6).

Fig 6
Fig 6:
Treatment with noggin in vivo may decrease the incidence of experimental metastases.


The authors thank Ms. Anne Olshanski, Dr. Bridget Deasy, Dr. Jonathan Pollett, and Ms. Marla Holderby for their valuable assistance in the compilation of these data. We also thank Lee Helman, MD, and Chand Khanna, DVM, PhD, at the National Cancer Institute for their generous gift of the K7M2 and K12 cells. This work was supported in part by the Henry J. Mankin Endowed Chair, the William F. and Jean W. Donaldson Endowed Chair, the Hirtzel Foundation, and the National Institutes of Health (R01-DE13 420).


1. Arihiro K, Inai K. Expression of CD31, Met/hepatocyte growth factor receptor and bone morphogenetic protein in bone metastasis of osteosarcoma. Pathol Int. 2001;51:100-106.
2. Audio Visual Interleaved (AVI) Constuctor. Software available at: Accessed January, 2006.
3. Bacci G, Briccoli A, Mercuri M, Ferrari S, Bertoni F, Gasbarrini A, Fabbri N, Cesari M, Forni C, Campanacci M. Osteosarcoma of the extremities with synchronous lung metastases: long-term results in 44 patients treated with neoadjuvant chemotherapy. J Chemother. 1998;10:69-76.
4. Balemans W, Van Hul W. Extracellular regulation of BMP signaling in vertebrates: a cocktail of modulators. Dev Biol. 2002;250: 231-250.
5. Bielack SS, Kempf-Bielack B, Delling G, Exner GU, Flege S, Helmke K, Kotz R, Salzer-Kuntschik M, Werner M,Winkelmann W. Zoubek A, Jurgens H, Winkler K. Prognostic factors in high-grade osteosarcoma of the extremities or trunk: an analysis of 1,702 patients treated on neoadjuvant cooperative osteosarcoma study group protocols. J Clin Oncol. 2002;20:776-790.
6. Biosystems A. User Bulletin #2 ABI Prism 7700 Sequence Detection System: Applied Biosystems 2001. Available at:
7. Bruland OS, Pihl A. On the current management of osteosarcoma: a critical evaluation and a proposal for a modified treatment strategy. Eur J Cancer. 1997;33:1725-1731.
8. Cross MJ, Dixelius J, Matsumoto T, Claesson-Welsh L. VEGF-receptor signal transduction. Trends Biochem Sci. 2003;28: 488-494.
9. Deasy BM, Jankowski RJ, Payne TR, Coa B, Goff JP, Greenberger JS, Huard J. Modeling stem cell population growth: incorporating terms for proliferative heterogeneity. Stem Cells. 2003;21: 536-545.
10. Estlin EJ, Veal GJ. Clinical and cellular pharmacology in relation to solid tumours of childhood. Cancer Treat Rev. 2003;29: 253-273.
11. Graepler F, Verbeek B, Graeter T, Smirnow I, Kong HL, Schuppan D, Bauer M, Vonthein R, Gregor M, Lauer UM. Combined endostatin/sFlt-1 antiangiogenic gene therapy is highly effective in a rat model of HCC. Hepatology. 2005;41:879-886.
12. Groppe J, Greenwald J, Wiater E, Rodriguez-Leon J, Economides AN, Kwiatkowski W, Affolter M, Vale WW, Belmonte JC, Choe S. Structural basis of BMP signalling inhibition by the cystine knot protein Noggin. Nature. 2002;420:636-642.
13. Hannallah D, Peng H, Young B, Usas A, Gearhart B, Huard J. Retroviral delivery of Noggin inhibits the formation of heterotopic ossification induced by BMP-4, demineralized bone matrix, and trauma in an animal model. J Bone Joint Surg Am. 2004;86: 80-91.
14. Jadlowiec J, Koch H, Zhang X, Campbell PG, Seyedain M, Sfeir C. Phosphophoryn regulates the gene expression and differentiation of NIH3T3, MC3T3-E1, and human mesenchymal stem cells via the integrin/MAPK signaling pathway. J Biol Chem. 2004;279: 53323-53330.
15. Kaya M, Wada T, Kawaguchi S, Nagoya S, Yamashita T, Abe Y, Hiraga H, Isu K, Shindoh M, Higashino F, Okada F, Tada M, Yamawaki S, Ishii S. Increased pretherapeutic serum vascular endothelial growth factor in patients with early clinical relapse of osteosarcoma. Br J Cancer. 2002;86:864-869.
16. Kearney JB, Kappas NC, Ellerstrom C, DiPaola FW, Bautch VL. The VEGF receptor flt-1 (VEGFR-1) is a positive modulator of vascular sprout formation and branching morphogenesis. Blood. 2004;103:4527-4535.
17. Khanna C, Khan J, Nguyen P, Prehn J, Caylor J, Yeung C, Trepel J, Meltzer P, Helman L. Metastasis-associated differences in gene expression in a murine model of osteosarcoma. Cancer Res. 2001;61:3750-3759.
18. Khanna C, Prehn J, Yeung C, Caylor J, Tsokos M, Helman L. An orthotopic model of murine osteosarcoma with clonally related variants differing in pulmonary metastatic potential. Clin Exp Metastasis. 2000;18:261-271.
19. Marina N, Gebhardt M, Teot L, Gorlick R. Biology and therapeutic advances for pediatric osteosarcoma. Oncologist. 2004;9: 422-441.
20. McGough RL, Lin C, Meitner P, Aswad BI, Terek RM. Angiogenic cytokines in cartilage tumors. Clin Orthop Relat Res. 2002;397: 62-69.
21. Mehdi R, Shimizu T, Yoshimura Y, Gomyo H, Takaoka K. Expression of bone morphogenetic protein and its receptors in osteosarcoma and malignant fibrous histiocytoma. Jpn J Clin Oncol. 2000;30:272-275.
22. Meyer JS, Mackenzie W. Malignant bone tumors and limbsalvage surgery in children. Pediatr Radiol. 2004;34:606-613.
23. Meyers PA. High-dose therapy with autologous stem cell rescue for pediatric sarcomas. Curr Opin Oncol. 2004;16:120-125.
24. National Center for Biotechnology Information (NCBI). Nucleotide-Nucleotide BLAST. Available at Accessed in December 2004.
25. Ribatti D. The crucial role of vascular permeability factor/vascular endothelial growth factor in angiogenesis: a historical review. Br J Haematol. 2005;128:303-309.
26. Simon MA, Springfield D. Surgery for Bone and Soft-Tissue Tumors. Philadelphia: Lippincott-Raven; 1998.
27. Sulzbacher I, Birner P, Trieb K, Pichlbauer E, Lang S. The expression of bone morphogenetic proteins in osteosarcoma and its relevance as a prognostic parameter. J Clin Pathol. 2002;55: 381-385.
28. Tammela T, Enholm B, Alitalo K, Paavonen K. The biology of vascular endothelial growth factors. Cardiovasc Res. 2005;65: 550-563.
29. Unni KK, ed. Dahlin's Bone Tumors: General Aspects and Data on 11,087 Cases. Philadelphia: Lippincott-Raven; 1996.
30. Urist MR. Bone: formation by autoinduction. Science. 1965;150: 893-899.
31. Urist MR, Grant TT, Lindholm TS, Mirra JM, Hirano H, Finerman GA. Induction of new-bone formation in the host bed by human bone-tumor transplants in athymic nude mice. J Bone Joint Surg Am. 1979;61:1207-1216.
32. Whelan JS. Osteosarcoma. Eur J Cancer. 1997;33:1611-1618. discussion 1618-1619.
33. Yoshikawa H, Rettig WJ, Takaoka K, Alderman E, Rup B, Rosen V, Wozney JM, Lane JM, Huvos AG, Garin-Chesa P. Expression of bone morphogenetic proteins in human osteosarcoma. Immunohistochemical detection with monoclonal antibody. Cancer. 1994;73: 85-91.
34. Yoshikawa H, Takaoka K, Masuhara K, Ono K, Sakamoto Y. Prognostic significance of bone morphogenetic activity in osteosarcoma tissue. Cancer. 1988;61:569-573.
35. Zimmerman LB, De Jesus-Escobar JM, Harland RM. The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell. 1996;86:599-606.
© 2006 Lippincott Williams & Wilkins, Inc.