Insulinlike Growth Factor I: A Potent Mitogen for Human Osteogenic Sarcoma Michael N. Pollak, * Constantine Polychronakos, Martine Richard

Insulinlike growth factor I (IGF-I), also known as somatomedin C, is a 76-amino acid mitogen known to stimulate the proliferation of many tissues (/). The binding of IGF-I to specific cell-surface receptor molecules is the first step in the mitogenic signal transduction pathway leading to IGF-I-stimulated cellular proliferation. Like other mitogen receptors, the IGF-I receptor consists of an extracellular ligand-binding domain, a transmembrane region, and an intracellular tyrosine-specific protein kinase domain (2). In vivo, serum IGF-I is dependent on growth hormone levels.

Vol. 82, No. 4, February 21, 1990

Materials and Methods Cells, growth factors, and antibodies. MG-63 immortalized human osteosarcoma cells (8), recombinant human IGF-I, insulin, and growth hormone were obtained from the American Type Culture Collection (Rockville, MD), Amersham Corp. (Arlington Heights, IL), Eli Lilly & Co. (Indianapolis, IN), and Sigma Chemical Co. (St. Louis, MO), respectively. The a-IR3 blocking antibody against the IGF-I receptor (9) was provided by S. Jacobs. Tissue culture and membrane preparation. MG-63 cells were routinely cultured in RPMI-1640 medium supplemented with 10% fetal calf serum. A plasma membrane-enriched subcellular fraction was prepared from primary tumor tissue (unfixed and flash-frozen) and 109 cultured MG-63 cells as previously described (10). For growth curves, cells were plated in quadruplicate in 2.5-cm 2 dishes in medium with 10% fetal calf serum, which was changed after 24 hours to serum-free media with various concentrations of IGF-I. • After 6 days, the cells were trypsinized and counted with a hemacytometer. Binding studies. Aliquots of the plasma membrane-enriched subcellular fractions were incubated at 4 °C for 20 hours with labeled IGF-I and varying concentrations of unlabeled growth factors as previously described (10). Thymidine incorporation. We plated 3 X 104 cells in 2.5-cm2 wells with medium containing serum. After 24 hours, the medium was changed to

serum-free medium, and the cells were cultured for an additional 24 hours. Incubation was performed with various concentrations of IGF-I under serumfree conditions. After 22 hours, 1 /zCi of tritiated thymidine/mL was added to each well. Two hours later, cells were washed in phosphate-buffered saline and precipitated in 10% trichloroacetic acid (TCA). TCA-insoluble material was collected by centrifugation and washed by repeated centrifugation in TCA. The final TCA-insoluble pellet was dissolved in NaOH and neutralized with HC1. We measured radioactivity using a liquid scintillation counter, and the numbers of cells were determined in replicate cultures. Rates of thymidine incorporation were calculated as disintegrations per minute per cell over 2 hours, and values were expressed as percentages of control values. Experiments were performed in quadruplicate. Affinity labeling. Plasma membraneenriched subcellular fractions were prepared as previously described (10) from human placenta, cultured cells, or primary tumor tissue and incubated for 24 hours at 4 °C with radiolabeled IGF-I (20,000 cpm) in the presence or absence of excess unlabeled IGF-I. Electrophoresis was performed on a 7.5% polyacrylamide gradient gel after cross-linking with disuccinimidyl suberate, solubilizing with sodium dodecyl sulfate, and reduction in 100 mM 1,4-dithiothreitol ( / / ) .

Results The results of binding studies on MG-63 osteosarcoma cells and primary osteogenic sarcoma cells are shown in figures 1 and 2, respectively. In each figure, panel A illustrates compe-

Downloaded from http://jnci.oxfordjournals.org/ at University of California, Santa Cruz on November 20, 2014

Insulinlike growth factor I (IGF-I) is among the peptide mitogens that play key roles in the regulation of normal skeletal growth. To investigate the possibility that certain skeletal neoplasms retain a sensitivity to mitogenic stimulation by IGF-I, we studied the effects of this growth factor on human osteosarcoma. Competitive-binding assays and affinity-labeling experiments on membranes prepared from MG63 immortalized human osteosarcoma cells and primary human osteogenic sarcoma cells demonstrate the presence of specific IGF-I receptors. Furthermore, we show that IGF-I is a potent stimulator of proliferation of MG-63 cells in vitro and is active at concentrations as low as 10~10 M. A blocking antibody against the IGF-I receptor (a-IR3) significantly reduces IGF-I-stimulated proliferation in a dose-dependent manner. These results are consistent with the hypothesis that at least a subset of human osteogenic sarcomas are responsive to IGF-I and indicate that it may be possible to exploit this responsiveness therapeutically. [J Natl Cancer Inst 82:301-305, 1990]

Although the liver is a major site of growth hormone-dependent IGF-I synthesis, recent work indicates that a variety of normal and neoplastic tissues, including sarcomas, may also produce IGF-I {3-5). Thus IGF-I may act as a regulator of cellular proliferation via autocrine or paracrine as well as endocrine mechanisms. In view of evidence that IGF-I is a mitogen for normal osteoprogenitor cells (6,7), we undertook the present studies to determine the degree to which osteogenic sarcoma might retain a dependence on this mitogen for optimum proliferation.

Received August 17, 1989; revised October 31, 1989; accepted November 3, 1989. Supported by a grant from the Cancer Research Society, Montreal, Canada. Departments of Medicine, Oncology, and Pediatrics, McGill University, Montreal, Canada. We thank Dr. Steven Jacobs for the kind gift of the or-IR3 antibody. Correspondence to: Michael N. Pollak, M.D., Lady Davis Research Institute of the Jewish General Hospital, 3755 Cote St. Catherine Rd., Montreal, PQ, Canada H3T 1E2.

REPORTS 301

120-,

A

tal Bin ding

100-

\

80 -

\ \

60 40 -

y \

200-

i

.1

1

i

10

100

i

i

1000

10000

B 604020

0.03

0.05

0.07

Bound (ng / 40 ug protein)

[alR I ng/ml

Figure 1. Binding of IGF-I to membranes prepared from human MG-63 osteosarcoma cells. A: Competition of binding of radiolabeled IGF-I with unlabeled IGF-I ( • ) and insulin ( • ) . B: Scatchard plot of binding data. C: Displacement of bound IGF-I by increases in concentration of a-IR3 antibody, a blocking antibody against the IGF-I receptor.

tition of radiolabeled IGF-I with unlabeled IGF-I and insulin for IGF-I binding sites; panel B is a Scatchard plot of the binding data; and panel C illustrates displacement of bound radiolabeled IGF-I from binding sites by increases in the concentration of a-IR3 antibody (9), a blocking antibody against the IGF-I receptor. Under our experimental conditions, we observed 16.4% and 16.0% specific binding of IGF-I to membranes prepared from MG-63 cells and primary osteogenic sarcoma cells, respectively, while human placenta, a tissue known 302

to have a particularly high concentration of IGF-I receptors (72), showed 25% specific binding. Scatchard analysis revealed equilibrium affinity constants (Kd) of 2.9 and 2.5 nM for MG-63 cells and primary osteogenic sarcoma cells, respectively. The calculated binding capacity for IGF-I per milligram of protein is 0.17 pmol for primary osteogenic sarcoma cells and 0.21 pmol for MG-63 cells. As the a-IR3 antibody is specific for IGF-I receptors, the data suggest that IGF-I binds to these receptors, rather than to receptors for insulin or IGF-II.

Discussion Our data suggest that at least a subset of human osteogenic sarcomas are IGF-I receptor positive and can be stimulated to proliferate by IGF-I. It is not possible to directly extrapolate our in vitro dose-response data to in vivo or clinical situations, partly because the nature and concentration of IGF-I binding proteins, which influence the bioactivity of somatomedins, are not identical in vitro and in vivo (13,14). However, the fact that IGF-I was acJournal of the National Cancer Institute

Downloaded from http://jnci.oxfordjournals.org/ at University of California, Santa Cruz on November 20, 2014

ng/ml

The similarity of binding data from the primary osteogenic sarcoma cells and the cultured MG-63 osteosarcoma cells indicates that in vitro proliferation of the latter cells may, in certain respects, reflect.the in vivo behavior of the former cells. In similar studies with radiolabeled growth hormone, we did not observe significant binding. Chemical cross-linking of 125I-labeled IGF-I to binding sites in membranes derived from MG-63 cells and primary osteogenic sarcoma cells and analysis by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (fig. 3) showed labeled proteins at the 130kilodalton band in a position similar to that seen in control membranes from human placenta. Labeling was totally abolished by the presence of excess unlabeled IGF-I, a finding consistent with the presence of specific IGF-I receptors. Some labeling was observed at the 260-kilodalton band. As this band is also abolished by excess cold IGF-I, it likely represents a receptor dimer. Figure 4A shows the stimulatory effect of IGF-I on the in vitro proliferation of MG-63 osteosarcoma cells, as measured by the number of cells. The number of cells at day 6 in the media containing IGF-I was significantly greater than in control media (P < .01). Figure 4B shows dose-response data for the same cell line incubated at various IGF-I concentrations, as determined by thymidine incorporation. Figure 4C illustrates the inhibition of IGF-I-stimulated proliferation of MG63 cells by the a-IR3 antibody. These data provide further support for the hypothesis that IGF-I stimulates proliferation of MG-63 cells by binding with IGF-I receptors.

60-

3

40-

20-

0

10

.1

100

1000

10000

ng/ml

Our results obtained with human tissues are consistent with previous data from an in vivo rat chondrosarcoma model (20). In this system, the effect of IGF-I was not studied directly, but tumor growth was greatly accelerated in the presence of growth hormonedependent serum factors. The responsiveness of osteogenic sarcoma to IGF-I may have clinical relevance, given that the peak incidence of osteogenic sarcoma occurs in adolescence, coincident with the pubertal peak in IGF-I levels. In view of our data, it is reasonable to speculate that lowering systemic IGF-I levels might reduce the rate of proliferation of osteogenic sarcomas in patients.

B

0.03

0.05

0.07

Bound (ng/40|xg protein)

[alR

l\iq/m\

3 Figure 2. Binding of IGF-I to membranes prepared from human primary osteogenic sarcoma cells. A: Competition of binding of radiolabeled IGF-I with unlabeled IGF-I ( • ) and insulin (D). B: Scatchard plot of binding data. C: Displacement of bound IGF-I by increasing concentrations of a-IR3 antibody, a blocking antibody against the IGF-I receptor.

a

200 —

tive under our experimental conditions at nanomolar concentrations indicates that normal serum IGF-I levels may be sufficient to stimulate the proliferation of osteogenic sarcomas in patients. The degree of stimulation of proliferation we observed (approximately sevenfold) is high compared with that observed in other IGF-I-responsive systems (1,11). Our data do not suggest that this responsiveness is related to an unusually high density of IGF-I receptors, although a previous study showed an association between hypersensitivity to epidermal growth factor and a Vol. 82, No. 4, February 21, 1990

high density of receptors for this growth factor in certain cell lines (15). We have considered the possibility that a derangement in the IGF-I receptor gene of MG-63 cells might result in abnormally high IGF-I-dependent tyrosine kinase activity, but studies to date have not shown such aberrations, at least in terms of phosphorylation of an artificial polytyrosine substrate (data not shown). It is possible, therefore, that the MG-63 cells studied were rendered unusually responsive to IGF-I by alterations in steps in the mitogenic signal transduction pathway that follow

b

c

d

e

f

1 I

Downloaded from http://jnci.oxfordjournals.org/ at University of California, Santa Cruz on November 20, 2014

receptor binding. As osteogenic sarcomas have been observed to have a homozygous deletion at the retinoblastoma locus (76), the responsiveness to IGF-I exhibited by MG-63 cells may be related to a derangement of negative growth controls, as discussed by Weinberg (17). It has been reported that platelet-derived growth factor is not a mitogen for MG-63 cells (18), despite the fact that these cells have receptors for this growth factor (19). The responsiveness of MG-63 cells to IGF-I stimulation is therefore specific rather than a reflection of general hypersensitivity to all mitogenic signals.

100-1

mm 116 —

Figure 3. Affinity labeling of IGF-I receptors of human osteogenic sarcoma cells. Membranes were incubated with labeled IGF-I in presence (lanes b, d, and 0 or absence (lanes a, c, and e) of excess unlabeled IGF-I. Cross-linking was carried out, and electrophoresis was performed on 7.5% polyacrylamide gel in presence of 2% sodium dodecyl sulfate. Lanes a and b: human placenta. Lanes c and d: MG-63 osteosarcoma cells. Lanes e and f: primary human osteogenic sarcoma cells.

REPORTS

303

•

20-

References

£

0

14"o * |

12

" 108-

4-

Control

20 0

7001

e

I •5 3?

1

2

3 days

4

5

6

E 100-,

B

600 80-

500-

•5 88

400300

60-

200-

I

100-

(/) 10 -12

10 -10

—1 10"8

MIGF1

0

0.5

Figure 4. Effect of IGF-I on proliferation of human MG-63 osteosarcoma cells. A: Effect of IGF-I on proliferation over 6 days, as measured by numbers of cells. B: Dose-response curve showing thymidine incorporation. C: Antiproliferative effect of IGF-I receptor blockade by a-IR3 antibody on human MG-63 osteosarcoma cells, as determined by thymidine incorporation. Cells were allowed to proliferate in serum-free media with IGF-I (10"' M), with varying amounts of receptor-blocking antibody.

(2)

304

required to determine what proportion of sarcomas have IGF-I receptors but no autocrine mechanism. Sarcomas with these characteristics might respond to treatments that lower IGF-I levels. The therapeutic potential of IGF-Ilowering strategies is supported by a significant decrease in the in vivo proliferation of experimental sarcomas achieved in two studies by treatment with somatostatin (23,24). As the sarcoma used in these experiments was somatostatin receptor negative, a direct antiproliferative effect of somatostatin is ruled out, and in view of

SCHMID

C,

FROESCH

ER:

Bio-

ULLRICH

A,

GRAY

A,

TAM

A,

ET AL:

Insulin-like growth factor I receptor primary structure: Comparison with insulin receptor suggests structural determinants that define functional specificity. EMBO J 5:2503-2512, 1986 (J)

BLATT J, WHITE C, DIENES S, ET AL:

Pro-

duction of an insulin-like growth factor by osteosarcoma. Biochem Biophys Res Commun 123:373-376, 1984 (4)

Apart from hypophysectomy, there are a variety of nonsurgical measures to lower IGF-I levels or block IGF-I action. These measures include treatment with long-acting somatostatin analogs and/or antagonists to growth hormonereleasing factor (27,22) and development of competitive antagonists to the IGF-I receptor. Lowering of IGF-I levels cannot be expected to slow the proliferation of sarcomas that are IGF-I receptor negative or those that produce significant quantities of IGF-I in an uncontrolled (growth hormone-independent) autocrine fashion. Further studies are

J,

logical and immunological properties of insulin-like growth factors (IGF) I and II. Clin Endocrinol Metab 13:3-30, 1984

1

[alR 3 J\

ZAPF

SCHLECTER NL,

RUSSEL SM, SPENCER

EM,

ET AL: Evidence suggesting that the direct growth-promoting effect of growth hormone on cartilage in vivo is mediated by local production of somatomedin. Proc Natl Acad Sci USA 83:7932-7935, 1986 (5)

HAN VK,

D'ERCOLE AJ,

LUND PK:

Cel-

lular localization of somatomedin messenger RNA in the human fetus. Science 236:193-195, 1987 (6)

CANELIS E, MCCARTHY T, CENTRELLA

M:

Growth factors and the regulation of bone remodeling. J Clin Invest 81:277-281, 1988 (7)

HOCK J, CENTRELLA M, CANALIS E: Insulin-

like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology 122:254-260, 1988 (8)

BILLIAU A, EDY VG,

HEREMANS H, ET AL:

Human interferon: Mass production in a newly established cell line, MG-63. Antimicrob Agents Chemother 12:11-15, 1977 (9)

JACOBS S, COOK S, SVOBODA ME, ET AL: In-

teraction of the monoclonal antibodies alpha IR1 and alpha IR3 with insulin and somatomedin-C receptors. Endocrinology 118:223-228, 1986

Journal of the National Cancer Institute

Downloaded from http://jnci.oxfordjournals.org/ at University of California, Santa Cruz on November 20, 2014

40-

our findings, an IGF-I-lowering mechanism of action is likely. Measures that reduce IGF-I stimulation of tumor proliferation are not tumoricidal. Nevertheless, given the precedent provided by the efficacy of hormonal treatments of steroid-dependent prostate and breast cancers, the potential role of IGF-I-lowering manipulations as welltolerated and effective palliative treatment deserves further study. A separate implication of our results is the possibility that the efficacy of cytotoxic chemotherapy of osteogenic sarcoma might be enhanced by the use of IGF-I in mitogenic stimulation of target cells in a manner analogous to the use of estrogens prior to chemotherapy for breast cancer (25). This synchronization strategy has been attempted clinically only with mitogenic steroids. As peptide growth factors are often more potent mitogens than steroids, they may be more effective in recruiting malignant cells into S phase prior to cytotoxic treatment.

16-i

(10)

POLLAK M, PERDUE J, MARGOLESE R, ET AL:

Presence of somatomedin receptors on primary human breast and colon carcinomas. Cancer Lett 38:223-230, 1987 (11)

POLLAK M, POLYCHRONAKOS C, YOUSEFI

S, ET AL: Characterization of insulin-like growth factor I (1GF-I) receptors of human breast cancer cells. Biochem Biophys Res Commun 154:326-331, 1988 (12) MASSAGUE J, CZECH M: The subunit struc-

tures of two distinct receptors for insulinlike growth factors I and II and their relationship to the insulin receptor. J Biol Chem 257:5038-5045, 1982 (13)

ELGIN RG, BUSBY WH, CLEMMONS DR: An

IGF binding protein enhances the biologic response to IGF-I. Proc Natl Acad Sci USA 84:3254-3258, 1987

( / 5 ) FILMUS J, TRENT J, POLLAK M , E T A L : Epider-

mal growth factor receptor gene-amplified MDA-468 breast cancer cell line and its nonamplified variants. Mol Cell Biol 7:251257, 1987 (16)

FRIEND S, BERNARDS R, ROGELJ S, ET AL:

A human DNA segment that predisposes to retinoblastoma and osteosarcoma. Nature 323:643-646, 1986 (17) WEINBERG R: Oncogenes, antioncogenes, and the molecular basis of multistep carcinogenesis. Cancer Res 49:3713-3721, 1989 (18) WOMER R, FRICK K, MITCHELL C, ET AL:

PDGF induces c-myc mRNA expression in MG-63 osteosarcoma cells but does not stimulate cell replication. J Cell Physiol 132:65-72, 1987 (19) GRAVES DT, ANTONIADES HN, WILLIAMS

SR, ET AL: Evidence for functional plateletderived growth factor receptors on MG-63 human osteogenic sarcoma cells. Cancer Res 44:2966-2970, 1984 (20) MCCUMBEE W, MCCARTY K, LEBOVITZ H:

Hormone responsiveness of a transplantable rat chondrosarcoma. II. Evidence for in vivo hormone dependence. Endocrinology 106:1930-1940, 1980 (21)

POLLAK M, POLYCHRONAKOS C, RICHARD M:

Somatostatin analogue SMS 201-995 reduces serum IGF-1 levels in patients with neoplasms potentially dependent on IGF-I. Anticancer Res 9:889-892, 1989 (22)

LUMPKIN MD, MULRONEY SE, HARAMATI A:

Inhibition of pulsatile growth hormone secretion (GH) and somatic growth in immature rats with a synthetic GH-releasing factor antagonist. Endocrinology 124:11541159, 1989 (23)

REDDING TW, SCHALLY AV: Inhibition of

growth of transplantable rat chondrosarcoma by analogs of hypothalamic hormones. Proc Natl Acad Sci USA 80:10781082, 1983 (24) REUBI J: A somatostatin analogue inhibits chondrosarcoma and insulinoma tumour growth. Acta Endocrinol (Copenh) 109:108-114, 1985 (25)

LIPPMAN ME, CASSIDY J, WESLEY M: A ran-

domized attempt to increase the efficacy of cytotoxic chemotherapy in metastatic breast cancer by normal synchronization. J Clin Oncol 2:28-36, 1984

Vol. 82, No. 4, February 21, 1990

The syndrome of inappropriate antidiuretic hormone (SIADH) secretion

Downloaded from http://jnci.oxfordjournals.org/ at University of California, Santa Cruz on November 20, 2014

(14) HINTZ RL: Plasma forms of somatomedin and the binding protein phenomenon. Clin Endocrinol Metab 13:31-42, 1984

commonly occurs in patients with small cell lung cancer (1-10). Experimental evidence {11-16) has suggested that Expression of the Atrial this syndrome is caused by the ectopic Natriuretic Factor Gene in production and secretion of arginir.e Small Cell Lung Cancer vasopressin (AVP). However, although measurement of plasma AVP in paTumors and Tumor Cell Lines tients with small cell lung cancer and SIADH may reveal normal AVP levels David P. Bliss, Jr., James F. (75), it has not been demonstrated in all Battey, R. Ilona Linnoila, tumors from such patients (15). When Michael J. Birrer, Adi F. tumor studies have failed to reveal AVP Gaidar, Bruce E. Johnson* production, mechanisms involving abnormal posterior pituitary secretion of AVP have been postulated (17). To circumvent difficulties with imHyponatremia in patients with small munohistochemical analysis of AVP in cell lung cancer can be caused by tumor specimens and the plasma AVP tumor production of arginine vasopeptide measurement, we studied AVP pressin (AVP) and result in the synmRNA expression in tumors and tudrome of inappropriate antidiuretic mor cell lines from patients with small hormone. In evaluating the expression cell lung cancer and either SIADH of AVP mRNA from tumor and tuor normal serum sodium values. In mor cell line specimens from five patients with small cell lung cancer and the initial experiments, not all patients hyponatremia (presumed to have the with SIADH expressed AVP mRNA, syndrome of inappropriate antidiuretic which corroborated the occasional abhormone), we found that the tumors sence of AVP peptide production preand tumor cell lines from two of these viously described. Small cell lung canfive patients expressed AVP mRNA. cer tumors or tumor cell lines, or both, The RNA samples from the three pa- are known to synthesize various peptients with undetectable AVP mRNA tides, such as adrenocorticotrophic horexpressed abundant atrial natriuretic factor (ANF) mRNA. Analysis of specimens from three patients with small cell lung cancer and normal serum sodium levels revealed no detectable Received October 20, 1989; revised November AVP mRNA expression, and samples 9, 1989; accepted November 15, 1989. from only one of these three patients' D. P. Bliss, Jr., J. F. Battey, R. I. Linnoila, specimens expressed detectable ANF M. J. Birrer, A. F. Gazdar, B. E. Johnson, mRNA. The AVP and ANF peptide lev- NCI-Navy Medical Oncology Branch, National els in lysate preparations of the tumor Cancer Institute and National Naval Medical cell lines from four of these patients Center, Bethesda, MD. Present address: D. P. Bliss, Jr., Department were tested by radioimmunoassay and of Surgery, Washington University School of confirmed the gene expression data. Medicine, St. Louis, MO. These studies demonstrate ectopic proM. J. Birrer, Department of Medicine, Uniduction of ANF mRNA in small cell formed Services University of the Health Scilung cancer specimens from patients ences, Bethesda, MD. We thank Ms. S. Jensen and Mr. J. Way for with this cancer and the syndrome technical assistance; Dr. H. Oie, Mr. E. Russell, of inappropriate antidiuretic hormone. and Ms. S. Stephenson for establishment of the These findings will be of particular small cell lung cancer cell lines used in this study; interest if future studies demonstrate and Drs. J. Minna, S. Wells, M. Kuehl, S. Sethat ectopic ANF production can cause gal, N. Seibel, and J. Brennan and Messrs. K. sodium abnormalities in patients with Nakahara and J. Way for helpful suggestions. We are also grateful to Drs. R. Scarborough and S. small cell lung cancer. [J Natl Cancer Hilliker of California Biotechnology, Inc., MounInst 82:305-310, 1990] tain View, CA, for providing the ANF probe phNFpUClA.713. * Correspondence to: Bruce E. Johnson, M.D., NCI-Navy Medical Oncology Branch, National Naval Medical Center, Bldg. 8, Rm. 5101, Bethesda, MD 20814.

REPORTS

305