J Cancer Res Clin Oncol (1992) 118:269-275

Ca~/cer~esearch

C ical @neology 9 Springer-Verlag 1992

Type I and Type II insulin-like growth factor receptors and their function in human Ewing's sarcoma cells F. van Valen 1, W. Winkelmann 2, and H. Jiirgens 1 1 Kinderklinik, Abt. f/Jr Pfidiatrische Hfimatologieund Onkologie, and 20rthop/idische Klinik, Heinrich-Heine-Universitfit, Moorenstrasse 5, D-4000 Dtisseldorf, Federal Republic of Germany Received 7 August 1991/Accepted26 November 1991 Summary. Binding studies using recombinant human 125I-labelled insulin-like growth factor I ([125I ] I G F - I ) revealed IGF-I receptors in three Ewing's sarcoma cell lines with Kd ranging from 7 4 x 1 0 - I 2 M to 1 0 0 x l 0 - 1 Z M and Bmax= 36-63 fmol/mg cell protein. [~25I]IGF-I binding was displaced by IGF-I, IGF-II and insulin with ICso values of 1.5 nM, 6.3 nM and 0.7 gM respectively. Recombinant human [125I]IGF-II radioligand-binding assays in the cell lines disclosed specific binding sites for IGF-II with K,~= (110-175) x 10-12 M and Bma x varying from 21 fmol/mg to 72 fmol/mg cell protein. Neither IGF-I nor insulin displaced [~2sI]IGF-II binding. IGF-I was found to increase basal glucose transport by maximally 1.5 times with ECso =0.9 nM IGF-I. The efficacy and potency of IGF-II on glucose uptake were comparable to those of IGF-I whereas insulin was ineffective. IGF-I and IGF-II also provoked stimulation of glycogen synthesis in Ewing's sarcoma cells. The maximal glycogenic response was reached at 0.01 gM IGF-I and 0.1 gM IGF-II, the ECso value being approximately 1 nM IGF-I and 2 nM IGF-II. Insulin did not significantly influence glycogen formation. IGF-I and IGF-II but not insulin increased DNA synthesis in Ewing's sarcoma cells. The maximal mitogenic response was obtained with 10 nM IGF-I or IGF-II with an ECs0 value of about 0.7 n M for both peptides, e-IR-3, a monoclonal antibody specific for the IGF type I receptor, effectively blocked IGF-I- and IGF-II-mediated metabolic responses. In conclusion, the data show that IGF-I and IGF-II induce rapid and longterm biological responses in Ewing's sarcoma cells exclusively through interaction with IGF type I receptors. Key words: Insulin-like growth factor receptor - Glucose transport - Glycogen synthesis - DNA synthesis - Ewing's sarcoma cells Offprint requests to: F. van Valen, Westf/ilischeWilhelms-Universit~it Miinster, Klinik und Poliklinik fiir Kinderheilkunde, P/idiatrische H/imatologieund Onkologie, Albert-Schweitzer-Strasse33, W4400 Miinster, Federal Republic of Germany

Introduction Insulin-like growth factors, IGF-I and IGF-II, are polypeptides homologous to insulin which are potent mitogens and cause metabolic changes in a variety of cell types through interaction with specific receptors. The receptor for IGF-I, structurally analogous to the insulin receptor, is a disulphide-bridged tetramer with a molecular mass of about 350 kDa, composed of two extracellular subunits of 130 kDa and two transmembrane/~ subunits with a molecular mass of 95 kDa and a ligand-stimulatable tyrosine-specific phosphotransferase (Massague and Czech 1982). Activation of the receptor tyrosine kinase is thought to be involved in the intracellular signalling of IGF-I and insulin (Czech 1989). The IGF-II receptor has been characterized as a transmembrane monomeric molecule of 270 kDa with no protein kinase activity (Morgan et al. 1987; Roth 1988). IGF-I and IGF-II share crossreactivity in their ability to bind to the type I and type II IGF receptor while insulin does not bind to the type II receptor (Sara and Hall 1990). Because of these overlapping binding specificities of the IGF receptors it is often difficult to establish which receptor may mediate a particular biological effect in response to IGF-I, IGF-II and insulin. Ewing's sarcoma is a malignant round, small-cell human tumour that originates in bone and extraskeletal tissues. The tumour arises predominantly in children and young adults and tends to metastasize rapidly. The histopathological differential diagnosis of Ewing's sarcoma includes small-cell osteosarcoma, rhabdomyosarcoma, retinoblastoma, neuroblastoma, and peripheral primitive neuroectodermal tumours (Schmidt and Harms 1990). The cellular origin of Ewing's sarcoma is a matter of dispute (Miettinen et al. 1982; Navas-Palacios et al. 1984). Several lines of argument suggest that Ewing's tumour may be neural in origin (Lipinski et al. 1987; Cavazzana et al. 1987; van Valen et al. 1988) and related to peripheral primitive neuroectodermal tumours (Whang-Peng et al. 1984; McKeon et al. 1988; Ambros et al. 1991).

270 The regulation o f Ewing's s a r c o m a cell proliferation and metabolism and the identity o f the receptors responsible for mediating mitogenic effects in response to hormones and factors in these turnout cells have remained undetermined. It was the aim o f the present study to characterize the I G F receptors on cultured Ewing's s a r c o m a cells and to assess their functional role in mediating the mitogenic and glucose metabolic actions o f I G F - I , I G F II and insulin in these cells. Preliminary aspects o f p a r t o f this w o r k have been presented (van Valen et al. 1990).

Materials and methods Recombinant human IGF-I (insulin-like growth factor I, somatomedin C) was purchased from Peninsula Laboratories (St. Helens, England) and recombinant human IGF-II from Sigma Chemical Co. (Miinchen, FRG). [i2sI]IGF-I (recombinant human 3-[125I]iodotyrosyl-IGF-I; 2118 Ci/mmol; 78 TBq/mmol), [azsI]IGF-II (recombinant human 3-[125I]iodotyrosyl-IGF-II; 2144Ci/mmol; 79 TBq/mmol), [methyl-aH]thymidine (25 Ci/mmol; 925 GBq/ mmol), D-[6-3H]glucose (0.5 Ci/mmol; 18.5 GBq/mmol) and 2deoxy-D-[1-1gC]glucose (59 mCi/mmol; 2.18 GBq/mmol) were obtained from Amersham Buchler (Braunschweig, FRG). The IGF-I receptor monoclonal antibody e-IR-3 was purchased from Dianova (Hamburg, FRG). Ewing's sarcoma cell lines (WE-68, VH-64, RM-82) were maintained in RPMI-1640 medium supplemented with 2 mM glutamine and 10% fetal calf serum and antibiotic/antimyeotic solution in a humidified atmosphere of 95% air/5% CO2. Cells were routinely subcultured in collagen-coated (5 ~g/cm2; Boehringer, Mannheim, FRG) 25-cm2 Falcon culture flasks every 5 days, 1 : 2, with trypsin (1 : 250; 0.05%)/EDTA (0.02%) in Puck's saline (Gibco, Grand Island, N.Y., USA). For studies on [125I]IGF-I and [~25I]IGF-II binding, glucose transport, glycogen metabolism and DNA synthesis, cells were propagated into collagen-coated (5 gg/cm2) 24-well Costar cluster plates. Receptor-binding studies were carried out as described recently (van Valen et al. 1989). In short, cells were incubated in Hanks' HEPES (pH 7.4; 300 l.tl) containing 1% human serum albumin (HH/HSA, binding buffer) and varying concentrations of [t25I]IGF-I or [i25I]IGF-II in the absence and presence of unlabelled peptide for 4 h at 20~ C. Preliminary time-course experiments showed that specific binding of [lzsI]IGF-I and [lZSI]IGF-II was maximal after 240 min at 20~ C and followed by a plateau up to 300 min. Specific binding measured at 20~ C was approximately sixfold higher than when measured at 37~ C, reaching a maximum after 60 rain and remaining elevated up to 300 min (data not shown). Cells were washed with a total of 1.5 ml (3 x 500 gl) ice-cold binding buffer, and the cell-associated radioactivity was determined in an LKB 1270 Rackgamma-II y counter after cell lysis with Lubrol-PX (1%, v/v)/HSA (1%) in H20 as described (van Valen et al. 1989). For competition experiments, cells were incubated with a fixed amount of [125I]IGF-I or [~2sIIIGF-II and varying concentrations of competing ligands. Glucose transport, using the non-metabolizable analogue 2deoxyglucose, was measured as reported previously (van Valen and Keck 1988 a). Briefly, cells at confluence were incubated at 37~ C for the indicated time periods with and without agonists in glucose-free HH/HSA (pH 7.4; 500 gl). During the last 15 min of this incubation 2-deoxy-D-[14C]glucose was added to the cultures at a final concentration of 0.4 gCi/ml. Incubations were stopped with ice-cold Hanks' HEPES (pH 7.4) containing 5.5 mM o-glucose, and the cells were lysed with 0.2 M NaOH and counted in a scintillation spectrometer. The incorporation of [all]glucose into glycogen and the [all]glycogen content in cell cultures were measured under conditions comparable to those previously reported (van Valen and Keck 1988 b).

Confluent cultures were incubated at 37~ C in HH/HSA (pH 7.4; 500 ~tl)containing 50 nM [all]glucose for 2 h, after which time 10 gt agonist solution was added, and the cells were incubated for the times indicated. Incubations were stopped by washing the cultures in HH/HSA buffer and cells were assayed for [3H]glycogen content by ethanol/trichloroacetic acid precipitation on paper discs as described (van Valen and Keek 1988 b). Experiments concerning [3H]thymidine incorporation as an index of DNA synthesis were performed in ceUs seeded in RPMI-1640 medium containing 5% fetal calf serum in 24-well plates. After 3 days, the medium was replaced by the same medium without serum and the test agents were than added to each well. After the appropriate incubation time (usually 20 h), medium was aspirated and replaced with 500 gl RPMI-1640 containing [3H]thymidine (2 gCi/ ml). After an incubation period of 5 h at 37~ C, labelling was stopped by removing the medium and washing the cells twice with HH/ HSA (pH 7.4) and three times with ice-cold 5% trichloroacetic acid. Trichloroacetic-acid-precipitable material was solubilized overnight with 0.2 M NaOH (250 ~l/well) and was then counted for radioactivity. The cell protein concentration was measured by the Bio-Rad protein assay (Bio-Rad, Miinchen, FRG) using HSA as standard. Data are expressed as means of triplicate determinations from at least three separate experiments. Statistical analysis was performed by Student's t-test. A value of P < 0.05 was considered statistically significant. Analysis of binding data was performed by the COMBICEPT 2000 CA receptor assay program (Canberra-Packard, Frankfurt, FRG) run on an IBM-PC computer. The apparent IC50 values for half-maximal inhibition of bound [125I]IGF-I and [lzsI]IGF-II by unlabelled IGF peptides and insulin, and the ECso values for half-maximal stimulation of glucose uptake, glycogen formation and thymidine incorporation by IGF peptides were calculated by linear regression analysis.

Results The specific binding o f [125I]IGF-I and [125I]IGF-II in the h u m a n Ewing's s a r c o m a cell line WE-68 is d e m o n strated in Fig. 1. [125I]IGF-I binding was displaced by I G F - I in a c o n c e n t r a t i o n - d e p e n d e n t fashion with ICs0 = 1.5 nM. I G F - I I was f o u n d to be only slightly less p o t e n t (4-fold) than I G F - I at inhibiting the binding o f [125I]IGF-I (ICso for I G F - I I = 6.3 nM), whereas insulin was 450 times less p o t e n t (IC5o for insulin--0.7 gM). [125I]IGF-II binding was displaced by I G F - I I with I C 5 0 = 2 . 7 nM, and I G F - I and insulin displayed no inhibitory effect up to 1 gM. The peptide specificity o f I G F - I and I G F - I I binding was established in the other two Ewing's s a r c o m a cell lines: V H - 6 4 and R M - 8 2 (data n o t shown). The specificity patterns were c o m p a r a b l e with that d e m o n s t r a t e d for Ewing's s a r c o m a cell line WE-68 in Fig. 1. A Scatchard plot o f saturation binding data in WE-68 cells, obtained with [~25I]IGF-I and [~2sI]IGF-II, is shown in Fig. 2. Linear slopes o f Scatchard plots were also determined in V H - 6 4 and R M - 8 2 cells (data n o t shown). The Kd values, Bin, x values and total cellular n u m b e r o f I G F receptors were estimated f r o m the Scatchard plots and are given in Table 1. The influence o f I G F - I , I G F - I I , and insulin on the uptake o f glucose in Ewing's s a r c o m a cells is reported in Table 2. D e p e n d i n g on the cell line employed, basal glucose u p t a k e was stimulated to different degrees in response to I G F - I and I G F - I I . Insulin did n o t alter the glucose u p t a k e rate significantly in the three Ewing's sarc o m a cell cultures studied. Since W E - 6 8 cells a p p e a r e d

271

100-

"'e~

v{,

O__(3__.__Z_

~A

0

p-

\

-100 T

(3 -Ts ~,

75-

"11

o z

r

z

50-

-50 z

(3

\

u,.

uJ

o~

(3

\

25o

I

"

-11

I

-10

-9

-8

-7

-6 PEPTIDE

I

I

-11

-10

CONCENTRATION

"25

t-

| -9

O -11 c~ o

O

-8

-7

-6

(M)

Fig. 1. Pharmacological specificity of 125i.labelled insulin-like growth factor I ([12sI]IGF.I) and [12sI]IGF.II binding in human Ewing's sarcoma WE-68 cells. Inhibition of specific p2sI]IGF-I binding by various concentrations of IGF-I (e), IGF-II ((3), and insulin (A) (left) and inhibition of specific [12sI]IGF-II binding by various concentrations of IGF-II ((3), IGF-I (o), and insulin (A) (right) expressed as a percentage of radioligand binding in the absence of unlabelled peptide. Data are the mean of triplicate determinations from three experiments Table 1. Dissociation constant (Ka), maximal binding capacity (Bin,x) and total receptors per cell (N) for insulin-like growth factor I (IGF-I) and IGF-II in human Ewing's sarcoma cells

0,20

B/F

0,15-

I

Ewing's sarcoma cell line

9

0.10

0.05-

WE-68 VH-64 RM-82

[]

9

18

27 36 BOUND Spec ( f tool/rag of protein )

Fig. 2. Scatchard plot of p/sI]IGF-I (m) and [125I]IGF-II (D) specific binding in human Ewing's sarcoma WE-68 cells. Data are the mean of triplicate determinations from four experiments. B, bound IGF; F, free IGF

most responsive to IGF-I (50 nM) stimulating glucose uptake by 54%, this cell line was used to characterize the time- and concentration-dependence of IGF-I, IGF-II, and inuslin on glucose transport. As shown in Fig. 3, IGF-I over the range 0.1 nM-0.1 gM enhanced the uptake of [14C]deoxyglucose in WE-68 cultures; a maximal increase (56% over basal) was provoked with 30 nM IGF-I with EC5o=0.9 nM. IGF-I stimulated the influx of [14C]deoxyglucose in a curvilinear fashion during the first 3 h, at which time the uptake rate reduced and a steady level was attained for at least another 3 h (Fig. 3, inset). IGF-II showed similar temporal characteristics and displayed similar potency (60% over basal at 30 nM IGF-II) and efficacy (ECso = 1 nM) as IGF-I. Insulin tested at up to 1 ~tM did not affect basal glucose transport, not even when cells were exposed for up to 6 h with insulin (Fig. 3). Under control incubation conditions the Ewing's sarcoma cell lines WE-68, VH-64 and RM-82 incorporated

[12sI]IGF-I Kd (pM)

Bmax

74 a 100 82

36 63 42

[12sI]IGF-II

N

Ka

(fmol/ rag) 3500 4358 3033

Bmax (fmol/ mg)

N

(pM)

110 175 174

21 72 28

1848 5111 2514

Data are expressed as the mean of three experiments Table 2. Effect oflGF-I, IGF-II, and insulin on glucose transport in human Ewing's sarcoma cell lines Ewing's sarcoma cell line WE-68 VH-64 RM-82

p4C]Deoxyglucose uptake rate (nm01/106 cells) Basal 24• 40• 62•

IGF-I

IGF-II

Insulin

(50 nM)

(50 nM)

(1 rtM)

37• 47• 84•

39• b(63) a 46• b(15) d 81+5 b (30) d

23• r 39• r 66+5 r

b(54) d b(18) d b (36) a

a Data are expressed as the mean _ SEM of three experiments b P