Vascular Endothelial Growth Factor Expression Is Higher in Differentiated Thyroid Cancer than in Normal or Benign Thyroid*
EUY Y. SOH, QUAN-YANG DUH, SAIF A. SOBHI, DAVE M. YOUNG,
HOWARD D. EPSTEIN, MARIWIL G. WONG, Y. KIT GARCIA, YOUNG D. MIN, RICHARD F. GROSSMAN, ALLAN E. SIPERSTEIN, AND ORLO H. CLARK Surgery (E.Y.S., Q-Y.D., S.A.B., D.M.Y., M.G.W., Y.D.M., R.F.G., A.E.S., O.H.C.) and Pathology (H.D.E.) Service, University of California San Francisco/Mount Zion Medical Center and Veterans Affairs Medical Center (Q-Y.D.), San Francisco, California, 94115; Genentech (K.G.), South San Francisco, California 94080-4990
ABSTRACT
Vascular endothelial growth factor (VEGF) is an angiogenic factor, and its expression has been rarely demonstrated in thyroid tumors.
We, therefore, investigated the expression of VEGF messenger RNA (mRNA) and production of VEGF protein in cell lines from human primary and metastatic follicular (FTC-133, FTC-236, and FTC-238), papillary (TPC-1), Hu¨ rthle cell (XTC-1), and medullary thyroid can- cers (MTC-1.1 and MTC-2.2), and in human thyroid tissues (papillary, follicular, medullary, and Hu¨ rthle cell cancers, follicular adenomas, and Graves’ thyroid tissue) by Northern blot, immunohistochemistry, and enzyme-linked immunosorbent assay (ELISA) studies. All thy- roid cell lines expressed a 4.2-kilobase VEGF mRNA. The VEGF mRNA levels were higher in the thyroid cancer cell lines than in primary cultures of normal thyroid cells, and higher in thyroid can- cers of follicular than those of parafollicular cell origin. The VEGF mRNA levels were similar in primary and metastatic thyroid tumors.
Immunohistochemical staining and Northern blot analysis of the cell lines correlated positively, thus thyroid cancer cell lines stained more intensely than normal thyroid cells and follicular tumor cells more intensely than parafollicular tumor cells. Again, no difference was noted in VEGF staining between primary and metastatic thyroid tumors. Deparafinized sections of papillary, follicular, and Hu¨ rthle cell cancers also stained much stronger than those of medullary thy- roid cancers, benign, or hyperplastic (Graves’ disease) thyroid tissue.
Thyroid cancer cell lines (XTC-1.TPC-1.FTC-133.MTC-1.1) also secreted more VEGF protein as measured by ELISA than did normal thyroid cells. VEGF secretion of cell lines derived from primary and metastatic thyroid tumors were similar. VEGF mRNA is therefore expressed, and VEGF protein is secreted by normal, hyperplastic, and neoplastic thyroid tissues. The higher levels of VEGF expression in differentiated thyroid cancers of follicular cell origin suggests a role in oncogenesis. (J Clin Endocrinol Metab 82: 3741–3747, 1997)
A
NGIOGENESIS is required for normal embryologic de- velopment and also appears to be involved in the abnormal growth of many tumors (1– 6). Although tumors 1–2 mm in size can be sustained by diffusion, further growth depends on angiogenesis.Vascular endothelial growth factor (VEGF) is unique among angiogenic factors because it is both mitogenic for vascular endothelial cells and is secreted by the cancer cells (7–10). The cancer cells, therefore, can stimulate the development of host blood vessels to bring more nutrients to support growth. VEGF is a 34- to 42-kilodalton heat- and acid-stable, dimeric, heparin-binding glycoprotein (11, 12). VEGF binds to membrane receptor tyrosine kinase (flt-1, KDR), and increases the proliferation of endothelial cells (13). VEGF also increases vascular permeability, caus- ing extravasation of plasma proteins and deposition of fibrin. This extravascular fibrin gel matrix then supports the ingrowth of new blood vessels (11, 14). Systemic ad-
ministration of anti-VEGF antibody has been found to inhibit the growth of xenografted tumor cells in nude mice (15, 16).
While investigating an in vivo invasion model, we ob- served extensive angiogenesis surrounding the human thy- roid cancer cells xenografted in nude mice. Because thyroid cancers are vascular tumors, and follicular cancer metasta- sizes via the blood vessel, we postulated that thyroid cancer cells produce and secrete VEGF, and that VEGF may stim- ulate thyroid cancer growth, invasion, and metastasis.
Material and Methods Cell lines and tissue preparation
We used human thyroid cell lines derived from follicular cancers (FTC-133, FTC-236, and FTC-238), kindly provided by Peter Goretzki, M.D., Du¨sseldorf, Germany, papillary cancer (TPC-1), kindly provided by Nobuo Satoh, Japan, Hu¨rthle cell cancer (XTC-1) developed in our laboratory, primary cultures of medullary thyroid cancer (MTC-1.1 and MTC-2.2), and normal thyroid tissue (NT 1.0). FTC-133, FTC-236, and FTC-238 were derived from primary tumor, metastatic lymph node, and pulmonary metastases from the same patient. MTC-1.1 was derived from the thyroid tumor and MTC-2.2 from a lymph node metastasis. The cells were obtained in accordance with approved human experimenta- tion protocols and maintained in standard incubating conditions (5%
CO2, 95% humidity, 37 C) in DMEM/F12 (Mediatech, Herndon, VA) supplemented with 10% FCS (Irvine Scientific, Santa Ana, CA), insulin (0.25 IU/mL; Sigma, St. Louis, MO), TSH (10 mIU/mL, Sigma), and Received March 22, 1996. Revision received January 8, 1997. Rerevi-
sion received May 18, 1997. Accepted July 10, 1997.
Address all correspondence and requests for reprints to: Quan-Yang Duh, Surgery Service, Veterans Affairs Medical Center, 4150 Clement Street, San Francisco, California 94121–11598.
* This work was supported in part by the Leonard Rosenman, M.D.
Fund, the Edwin H. Zeller Fund, and the Medical Research Service of the Veterans Affairs Medical Center.
Copyright © 1997 by The Endocrine Society
3741
antibiotics. A cell line from a patient with colon cancer (Colo-201 cell line), was obtained from the American Type Culture Collection (Rock- ville, MD) and grown in RPMI 1640 containing 20% FCS.
RNA extraction and Northern analysis
Total RNA was prepared from cultured cells using the RNA stat-60 (Tel-TestB, Friendswood, TX). For Northern analysis, RNA samples (15 mg) were size fractionated on 1% agarose gels containing 6% formal- dehyde and transferred to nitrocellulose membranes (Hybond-N; Am- ersham Life Sciences, Arlington Heights, IL) with low-vacuum (785 vacuum blotter; Bio-Rad, Richmond, CA), cross-linked with ultraviolet (UV) (UV cross-linker; Fisher Scientific) and then dried in a vacuum oven (80 C) for 2 h. The RNA was hybridized for 12–36 h at 42 C in 50%
deionized formamide, 4.73SSPE, 0.473Denhart’s solution, 0.1% SDS, and 10% dextran sulfate (17). The DNA probes (kindly provided by Dr.
Robert B. Jaffe, University of California, San Francisco, CA) were labeled using random primers labeling method with [a-32P]deoxycytidine triphosphate (Amersham) to a SA of 2–33108cpm/mg DNA. Typically, 2–33107cpm of32P-labeled probe was used for 70 cm2filter in 10 mL hybridization room temperature, 0.13 SSC/0.1% SDS for 20 min at 50 –55 C, and exposed to Kodak XAR-5 film (Eastman Kodak, Rochester, NY) at280 C with intensifying screens for 2–3 days. To control for the integrity and amount of messenger RNA (mRNA), the membranes were stripped in 0.1% SDS/0.013SSC for 10 min and reprobed forb-actin mRNA.
Immunohistochemistry
Immunohistochemistry studies were done both in cell lines and in the paraffin-embedded tissues. The thyroid cell lines were cultured on chamber slides (Nunc, Naperville, IL) for 2–3 days and fixed with cold acetone. The human thyroid tissues used for immunohistochemistry were initially fixed by 10% buffered formalin phosphate and embedded in paraffin. Tissues studied included Graves’ thyroid tissue, follicular adenomas, papillary cancers, follicular cancers, medullary cancers, and anaplastic cancers. Five-micrometer thick tissue sections were deparafinized and hydrated in PBS, incubated for 30 min in 1% hydro- gen peroxide to inhibit endogenous peroxidase, and incubated in 2%
goat serum in PBS for 20 min to block nonspecific binding. The sections were then incubated for 30 min at room temperature with a rabbit anti-VEGF polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:200 dilution (0.5mg/mL) using PBS and 1% BSA, and washed three times in PBS. The secondary antibody was biotinylated goat an- tirabbit IgG (1:300 dilution; Vector Labs., Burlingame, CA). The second- ary antibody was then detected using the avidin-peroxidase complex (Vector Labs.) for 30 min. The sections were stained by diaminobenzi- dine solution (Santa Cruz Biotechnology) for 5 min and counterstained with hematoxylene. For negative controls, we used all reagents except the primary anti-VEGF antibody.
Western blot
Cell lysates from the cell lines [XTC-1, TPC-1, FTC-133, FTC-236, FTC-238, and control human umbilical cord cells (HUV-EC-C cell lines)]
were prepared by washing cells once with PBS. Cell pellet(s) were obtained by centrifugation. PBS buffer was aspirated off, and 1 mL 6m Urea1b-mercaptoethanol was added. The cell lysate was sonicated for 1 min. Proteins were quantified using the Lowry protein assay method.
Two SDS-PAGE Gels (Biorad Ready-made 4 –15% gel) were prepared (one gel for transfer and the other for Coomassie blue staining). Ten to fifty micrograms of proteins were loaded onto each well using a 1:1 mix of protein volume and a 23loading buffer (Novex, San Diego, CA).
Samples were denatured at 90 C for 5 min and were quickly cooled on ice. Electrophoresis was carried out at 70 V for 3 h. One gel was stained with Coomassie brilliant solution (50% methanol, 0.05% Coomassie bril- liant blue R, 10% acetic acid, 40% water) for 4 h. It was destained with destaining solution (5% methanol, 7% acetic acid, 88% water) overnight.
The second gel was transferred onto a Hybond-N membrane (Amer- sham) for 3 h at 50 V using a Biorad transblot apparatus containing transfer buffer (39 mmglycine, 50 mmTris base, 20% methanol). After the overnight transfer, the membrane was dried for 5 min. The mem-
brane was blocked using 3% BSA/TBST for 1 h. The blocking reagent was replaced with the primary antibody (Santa Cruz Biotechnology, 1:100 in TBST buffer) and incubated for 2 h. The membrane was washed three times with TBST buffer for 5 min/wash. The TBST buffer was removed, and the membrane was incubated with secondary antibody (biotinylated antibody and/or HRP-labeled second antibody diluted in 3% goat serum/TBST) for 1 h. The membrane was then washed three times with TBST for 5 min/wash. The buffer was removed, and the membrane was incubated with streptavidin/HRP complex diluted in TBST for 1 h. The membrane was washed three times with TBST. The last TBST buffer wash was removed, and the membrane was incubated for 1 min in detection solution (1:1 vol of detection reagent 1 and detection reagent 2 from Amersham, Arlington Heights, IL). The de- tection solution was removed, and the membrane was wrapped in plastic wrap. The membrane was exposed for 15– 60 sec on Hyperfilm- ECL and developed.
Enzyme-linked immunosorbent assay (ELISA)
Cells (105) were cultured in 200mL MEM110% FCS for 7 days.
Aliquots of culture media were collected and assayed in ELISA. ELISA plates were coated with 2.5mg/mL monoclonal antibody (mAb) to VEGF (mAb 3.5F8; Genentech, South San Francisco, CA) in 50 mm carbonate buffer, pH 9.6, at 4 C overnight and blocked with 0.5% BSA in PBS. Standards (0.03–2 ng/mL recombinant V165, Genentech) and 3-fold serially diluted samples (initial dilution 1:5) in PBS contain- ing 0.5% BSA, 0.05% polysorbate 20, 0.25% 3-[(3-cholamidopropyl)- dimethyl ammoniol]-1-propane sulfate (Sigma), 0.2% bovineg-globulin (Sigma), 5 mmEDTA, and additional 0.35nNaCl were incubated on the plate for 2 h. Bound VEGF was detected using biotinylated mAb to VEGF (mAb 4.6.1, Genentech), followed by streptavidin peroxidase (Sigma) and 3,39,5959-tetramethyl benzidine (Kirgaard & Perry Labs.) as the substrate. Plates were washed between steps. Absorbance was read at 450 nm on Vmaxplate reader (Molecular Devices, Menlo Park, CA). The standard curve was fitted using a four-parameter nonlinear regression curve-fitting program (developed at Genentech). Data points that fell in the linear range of the standard curve were used for calculating the VEGF concentration in samples.
Statistical method
Statistical significance of ELISA data was determined by Student’st test.
Results Northern blots of thyroid cell lines
Cultured human thyroid cell lines expressed a 4.2-kilobase VEGF mRNA. The VEGF mRNA levels were higher in thy- roid cancer cell lines than in normal thyroid cells. Thyroid cancer cell lines of follicular cell origin (FTC-133, TPC-1, and XTC-1) expressed more VEGF mRNA than did thyroid cell lines of parafollicular cell line (MTC-1.1) (Fig. 1A). There were no differences in the VEGF mRNA signals between the cells derived from the primary thyroid tumors and metas- tases (FTC-133vs. FTC-236 or FTC-238, MTC-1.1vs. MTC-2.2, Fig. 1B).
Western blots of thyroid cell lines
Western blots showed production of VEGF121, VEGF165, and VEGF189splice variants by the XTC-1 (Hu¨rthle cell) and TPC-1 (papillary cancer) cell lines (Fig. 1C).
Immunohistochemical studies in thyroid cell lines and human thyroid tissue
The findings of immunohistochemical staining of the cell lines parallels the findings of the Northern blots (Fig. 2).
Thyroid cancer cell lines stained more positively than normal thyroid cells, and thyroid cancer cells of follicular cell origin stained stronger than thyroid cancer cells of parafollicular cell origin. Similar VEGF staining was again observed in primary and metastatic thyroid cancer cell lines. The deparafinized sections of human papillary thyroid cancer, follicular thyroid cancer, and Hu¨rthle cell thyroid cancer also stained strongly. Adjacent normal thyroid tissues stained only weakly compared with the cancers (Fig. 3). Benign thy- roid tumors (follicular adenoma), hyperplastic thyroid tissue (Graves’ thyroid), and medullary thyroid cancer also stained less strongly than did thyroid cancers of follicular cell origin.
Measurement of VEGF secretion by ELISA
All thyroid cells secreted VEGF in the conditioned me- dium as determined by ELISA (Fig. 4A). Thyroid cancer cell lines secreted more VEGF protein than did normal thyroid cells in primary culture. Among the cancer cell lines, XTC-1 secreted the most (39.1 ng/mL) and MTC-1.1 secreted the least (15.0 ng/mL) VEGF. There was no difference in VEGF secretion between the cell lines derived from the primary thyroid tumors and metastases (FTC-133vs. FTC-236 or FTC- 238, MTC-1.1,vs. MTC-2.2, Fig. 4B). Under the conditions in which these cells were grown, there was no appreciable cell proliferation during this period as measured by the MTT assay. Correction for cell proliferation was therefore not necessary.
Discussion
Many embryologic tissues and some cancers produce and secrete VEGF (6). Because VEGF is required for the growth of other solid tumors (18, 19), it is likely that it is also im- portant for the growth of thyroid cancers. We found that human thyroid cancers express more VEGF mRNA and pro- duce more VEGF protein than normal thyroid tissues.
Immunohistochemical studies also demonstrated that can- cer tissues stained more intensely than benign thyroid tu- mors and more intensely than adjacent normal tissues. The patterns of mRNA expression as well as immunohistochem- ical staining were not different between the primary thyroid tumors (FTC-133 and MTC-1.1) and respective metastases in lymph nodes (FTC-236 and MTC-2.2) or lung (FTC-238).
Our findings are consistent with other studies that found higher levels of VEGF mRNA in colon cancer and renal cell carcinoma than adjacent normal tissues and higher levels in tumorigenic cell lines (HT 1080 fibrosarcoma, MNNG HOS osteosarcoma) than in nontumorigenic cell lines.
Viglietto et al. (20) showed that thyroid tumors ex- pressed more VEGF mRNA and stained stronger with anti-VEGF antibody than normal thyroid tissue. Ours is
XTC-1 (Hu¨rthle cell cancer), FTC-236 (follicular cancer derived from lymph node metastasis), and FTC-238 (follicular cancer derived from lung metastasis). C, Western blot using polyclonal anti-VEGF anti- body that was developed from amino terminal epitope (residue 1–20) of human VEGF (A-20, Santa Cruz Biotech.) It shows three bands corresponding to VEGF121, VEGF165, and VEGF189splice variants for XTC-1 (Hu¨rthle cell), TPC-1 (papillary thyroid cancer), and HUV- EC-C (control, human umbilical cord cells).
FIG. 1. A, Northern blot analysis of VEGF mRNA in thyroid cancer cell lines. Transcripts of approximately 4.2 kilobases are present.
Thyroid cancer cell lines, except MTC-1.1 (medullary cancer), ex- pressed more VEGF mRNA than NT 1.0 (normal thyroid). B, Cell lines derived from primary tumor and cell lines derived from metastatic site did not differ in VEGF mRNA expression. Colo-201 served as a positive control. FTC-133 (follicular cancer), TPC-1 (papillary cancer),
FIG. 2. Immunohistochemical staining of thyroid cancer cell lines and thyroid tissue specimen with an antihuman VEGF antibody. Staining was detected in cytoplasm. Thyroid cancer cell lines (C, D, and E), except MTC-1.1 (F), showed stronger staining than NT 1.0 (B). Cell lines derived from primary tumor and cell lines derived from metastases (C vs. G1H, and F vs. I) did not differ in staining density. Note stronger staining in follicular thyroid cancer (K), papillary thyroid cancer (L), and Hu¨ rthle cell cancer (M) compared with medullary thyroid cancer (N), adjacent normal thyroid tissue of follicular thyroid cancer (J), and benign thyroid disease such as follicular adenoma (O) and Graves’ thyroid (P). A, Control of TPC-1. C, FTC-133. D, TPC-1. E, XTC-1. G, FTC-236. H, FTC-238. I, MTC-2.2.
the first study, however, to show that Hu¨rthle cell thyroid cancer and medullary thyroid cancer cells also express VEGF mRNA. Viglietto and co-workers found that the expression of VEGF seems to correlate with the aggres- siveness of the tumorsin vivoand their tumorigenic ability, and they showed that TPC-1 did not express VEGF. In contrast, we found TPC-1 cell line to express VEGF mRNA, to stain strongly with anti-VEGF antibody, and to secrete more VEGF into the condition media than the follicular thyroid cell lines FTC-133, FTC-236, and FTC-238. One may also have expected that the cell lines derived from metastases (FTC-236 and FTC-238) would have ex- pressed more VEGF than the cell line derived from the primary thyroid cancer (FTC-133) if VEGF expression cor- relates with the ability of the cells to metastasize, but we did not find any difference among these three cell lines derived from the same patient.
The concentrations of VEGF in the conditioned media from the thyroid cell lines were significantly higher (15.8 – 39.1 ng/mL) than the concentration in the ocular fluid (3.6 ng/mL) obtained from patients with active proliferative di- abetic retinopathy (21) and similar to the concentration in the conditioned media of G55 glioblastoma multiforme (41 ng/
mL) and SK-LMS-1 leiomyosarcoma (14 ng/mL) cell lines (15). It should be pointed out, however, that our results were obtained from a closed culture system in which the VEGF can accumulate. The ocular fluid samples, in contrast, were iso- lated and quantitatedin vivofrom a open system with as- sociated transport, clearance, metabolism, etc., which may account for lower values of VEGF. The high concentration of VEGF in the conditioned media suggests that VEGF may be more important for the growth and invasion of thyroid can- cers than in other cancers.
Follicular thyroid cancers are vascular tumors and metas- tasize hematogenously. The effect of VEGF in increasing vascular permeability may facilitate this tendency for hema- tological dissemination (23, 24). Small papillary thyroid can- cers are relatively common tumors but are rarely clinically significant because they usually do not grow appreciably or metastasize (25). Although it is possible that VEGF expres- sion may be increased in the more metastatic phenotype (20, 26), we observed no appreciable differences among cell lines derived from primary tumors and cell lines from metastatic sites from the same patient, or among tumors of differenti- ated histological types.
It is intriguing, however, that the only medullary cancer we studied produced less VEGF than the other thyroid can- cers. Medullary thyroid cancers are derived from the parafol- licular cells and are generally more aggressive than cancers derived from the follicular thyroid cells. It appears that other angiogenic factors are more important in supporting the growth and invasion of medullary thyroid cancers.
We have also recently found that VEGF expression is stim- ulated by TSH in thyroid cancer cell lines (27). In this study, the cell line that had the highest basal secretion of VEGF (XTC-1, Hu¨rthle cell cancer) was least stimulated by TSH, whereas the FTC lines (follicular cancer) and the TPC-1 (pap- illary cancer) were significantly stimulated by TSH and in- creased their VEGF production by 2- to 5-fold.
In conclusion, we demonstrated increased VEGF expres- sion and secretion in differentiated thyroid cancers of fol- licular cell origin. It is likely that VEGF is important for thyroid tumor growth and invasion. Antiangiogenic agents may thus be useful in treating patients with papillary and follicular thyroid cancers.
FIG. 3. Immunohistochemical staining of deparafinized section of a follicular thyroid cancer using a rabbit anti- VEGF polyclonal antibody (Santa Cruz Biotech.) showing that cancer tissue stained more intensely than adjacent normal thyroid tissue.
Acknowledgments
We thank K. Jin Kim, PhD and Gloria Meng, PhD for helpful dis- cussions about ELISA.
References
1. Folkman J. 1990 What is the evidence that tumors are angiogenesis dependent?
J Natl Cancer Inst. 82:4 – 6.
2. Risau W. 1990 Angiogenic growth factors. Prog Growth Factor Res. 2:71–79.
3. Bouck N. 1990 Tumor angiogenesis: the role of oncogenesis and tumor sup- pressor genes. Cancer Cells. 2:179 –185.
4. Folkman J, Watson K, Ingber D, Hanahan D. 1989 Induction of angiogenesis during the transition from hyperplasia to neoplasia. Nature. 339:58 – 61.
5. Folkman J, Shing Y. 1992 Angiogenesis. J Biol Chem. 267:10931–10934.
6. Shifren JL, Doldi N, Ferrara N, Nesiano S, Jaffe RB. 1994 In the human fetus, vascular endothelial growth factor is expressed in epithelial cells and mono- cytes, but not vascular endothelium: Implications for mode of action. J Clin Endocrinol Metab. 79:316 –322.
7. Conn G, Soderman DD, Schaeffer MT, Wile M, et al. 1990 Purification of a glycoprotein vascular endothelial cell mitogen from a rat glioma-derived cell line. Proc Natl Acad Sci USA. 87:1323–1327.
8. Leung DW, Cachianes G, Kuang W-J, Goeddel DV, Ferrara N. 1989 Vascular endothelial factor is a secreted angiogenic mitogen. Science. 246:1306 –1309.
9. Jakeman LB, Altar CA, Winer J, Bennett GL, Ferrara N. 1992 Binding sites for vascular endothelial growth factor are localized on endothelial cells in adult rat tissues. J Clin Invest. 89:244 –253.
10. Weinder K, Marme D, Weich H. 1992 AIDS-associated Kaposi’s sarcoma cells in culture express vascular endothelial growth factor. Biochem Biophys Res Commun. 183:1167–1174.
11. Senger DR, Connolly DT, Van De Water L, Feder J, Dvorak HF. 1990 Pu- rification and NH2-terminal amino acid sequence of guinea pig tumor-secreted vascular permeability factor. Cancer Res. 50:1774 –1778.
12. Yeo T-K, Senger DR, Dvorak HK, Freter L, Yeo K-T. 1991 Glycosylation is essential for efficient secretion but not for permeability enhancing activity of vascular permeability factor (vascular endothelial growth factor): Biochem Biophys Res Commu. 179:1568 –1575.
13. Brock TA, Dvorak HF, Senger DK. 1991 Tumor-secreted vascular permeabil- ity factor increases cytosolic Ca21and von Willebrand factor release in human endothelial cells. Am J Pathol. 138:213–221.
14. Nagy JA, Brown LF, Senger DR, et al. 1988 Pathogenesis of tumor stroma generation: a critical role of for leaky blood vessels and fibrin deposition.
Biochem Biophys Acta. 948:305–326.
15. Kim KJ Li B, Winer J, Armanini M, Gillett N, Philips HS, Ferrara N. 1993 FIG. 4. ELISA showing concentration
of VEGF in conditioned media of thy- roid cancer cell lines. A, All thyroid can- cer cell lines secreted more VEGF than normal thyroid cell. B, Cell lines de- rived from primary tumor and cell lines derived from metastatic site did not dif- fer in concentration of VEGF. Values are mean6SD.
Inhibition of vascular endothelial growth factor-induced angiogenesis sup- presses tumor growthin vivo. Nature. 362:841– 844.
16. Warren RS, Yuan H, Matli MR, Gillett NA, Ferrara N. 1995 Regulation by vascular endothelial growth factor of human colon cancer tumorigenesis in a mouse model of experimental liver metastasis. J Clin Invest. 95:1789 –1797.
17. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning: A laboratory manual, 2nd ed, Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
18. Dvorak HF, Rosen S. 1992 Vascular permeability factor mRNA and protein expression in human kidney. Kid Int. 42:1457–1461.
19. Plate KH Breier G, Weich HA, Risau W. 1992 Vascular endothelial growth factor is a potential tumor angiogenesis factor in human gliomas in vivo.
Nature. 359:845– 848.
20. Viglietto G, Maglione D, Rambaldi M, et al. 1995 Upregulation of vascular endothelial growth factor (VEGF) and down regulation of placental growth factor (PLGF) associated with malignancy in human thyroid tumors and cell lines. Oncogene. 11:1569 –1579.
21. Berse B, Brown LF, De Water LV, Dvorak HF, Senger DR. 1992 Vascular permeability factor (Vascular endothelial growth factor) gene is expressed differentially in normal tissues, macrophage, and tumors. Mol Bio Cell.
3:211–220.
22. Aiello II, Avery RL, Arrigg PG, et al. 1994 Vascular endothelial growth factor in ocular fluid of the patients with diabetic retinopathy and other retinal disease. N Eng J Med. 331:1480 –1510.
23. Asano M, Yukita A, Matsumoto T, Kondo S, Suzuki H. 1995 Inhibition of tumor growth and metastasis by an immunonutralizing monoclonal antibody to human vascular endothelial growth factor/vascular permeability factor121.
Cancer Res. 55:5296 –5301.
24. Takahashi Y, Kitadai Y, Bucana CD, Cleary KR, Ellis LM. 1995 Expression of vascular endothelial growth factor and its receptor, KDR, correlate with vascularity, metastasis, and proliferation of human colon cancer. Cancer Res.
55:3964 –3968.
25. Cady B, Rossi R. 1991 Surgery of the thyroid and parathyroid gland. In:
Silverman ML, ed. Pathology of the thyroid and parathyroid gland. 3rd ed.
Philadelphia: WB Saunders; 31– 44.
26. Folkman J, Hanahan D. 1991 Switch to the angiogenic phenotype during tumorigenesis. Princess Takamatsu Symposia. 22:337–347.
27. Soh EY, Al-Sobhi S, Wong M, et al. 1996 Thyroid stimulating hormone (TSH) promotes the secretion of vascular endothelial growth factor (VEGF) in thyroid cancer cell lines. Surgery. 120:944 –947.