Caterina P. Minniti$j, Elise C. Kohnll, Jeff H. Grubbll, William S. SlyII, Youngman Oh**, ..... unstimulated motility, were placed in the lower wells as chemoattrac-.
THEJOURNALOF BIOLOGICAL CHEMISTRY
Vol. 267, No. 13, Issue of May 5, pp. 9000-9004, 1992 Printed in U.S.A.
The Insulin-like Growth Factor I1 (IGF-II)/Mannose6-Phosphate Receptor Mediates IGF-11-induced Motility in Human Rhabdomyosarcoma Cells* (Received for publication, October 23, 1991)
Caterina P. Minniti$j, Elise C. Kohnll, Jeff H. Grubbll, WilliamS. SlyII, Youngman Oh**, Hermann L. Muller**$$, Ron G. Rosenfeld**,and Lee J. Helman$ From the $Molecular Genetics Section, Pediatric Branch and (Medicine Branch, NationalCancer Institute, National Institutes of Health, Bethesda, Maryland20892, the 11Department of Biochemistry, St. Louis University School of Medicine, St.Louis, Missouri 63104, and the **Stanford University Medical Center, Stanford, California 94305
Insulin-like growth factor-I1 (IGF-11)is an autocrine growth and motility factor for human rhabdomyosarcoma. It interacts with three different receptors: the IGF-I, the IGF-11, and the insulin receptor. A specific function of the IGF-I1 receptor in mediating IGF-I1 responses has not been defined. In this report we investigate the mechanism of IGF-11-mediated motility in rhabdomyosarcomacells. We demonstrate that IGF-I1 and [Leu2’]IGF-II, an analog selective for the IGF-I1 receptor, stimulate motility at concentrations in which they interact only with their own receptor. An antibody that blocks the IGF-I receptor does not inhibit either peptide activity, while an antibody specific for the IGF-I1 receptor suppresses the IGF-IIinduced motility. This antibody does not interfere with rhabdomyosarcoma cell proliferation. We conclude that in rhabdomyosarcomacells IGF-I1 stimulates two different responses mediated by distinct receptors: 1) a mitogenic response through the type I receptor and 2) a motility response through the type I1 receptor.
heterotetrameric structures and possess an intrinsic ligand-activated tyrosine kinase activity that appears to be necessary for cellular signal transmission (10). To date, the type I IGFreceptor has been shownto mediate most of the anabolic and mitogenic effects of IGF-I and IGFI1 (for review see Ref. 8). The type I1 IGF receptor binds preferentially to IGF-11, has 100-fold less affinity for IGF-I, and does not bind insulin (11, 12). Recently, this receptor and the cation-independent mannose 6-phosphate receptor (CIMPR) have been demonstrated to constitute a single multifunctional receptor (13-15). The IGF-II/CI-MPR is a transmembrane monomeric molecule that binds the two classes of ligands to different sites (16, 17) and has no demonstrable tyrosine kinase activity. It has been reported that theIGF-11/ CI-MPR is coupled to a Guizguanine nucleotide-binding protein and stimulation with IGF-I1 activates phosphoinositol metabolism and a calcium-permeable channel (18, 19). While its function in mediating the transport of lysosomal enzymes from the Golgi to prelysosomes and the endocytosis of Man-6-P-containing ligands (20) has been clearly established, attempts at defining the biological relevance of the binding of IGF-I1 tothe IGF-II/CI-MPR havebeen less The insulin-like growth factors IGF-I1 and IGF-I1 are two successful. It has been suggested that IGF-I1 binding to its single chain polypeptide hormones whose amino acid sequence receptor stimulates glycogen synthesis in human hepatoma and predicted three-dimensional structure are homologous to cells (21), stimulates the proliferation of K-562 cells (22), that of proinsulin (1-3). In recent years these factors have causes alkalinization of cells of the proximal tubular segment been shown to actby endocrine, autocrine, or paracrine path- of canine kidney by stimulating Na+/H+exchange across the ways and to play a role in the growth regulation of several brush border membrane (23), and stimulatesCa2+ influx and DNA synthesis in competent Balb/c 3T3 cells primed with human malignancies (4-7). Three major cell surface receptors bind thesepeptides with epidermal growth factor (16). We have demonstrated that IGF-I1 functions as an autodifferent affinities (8, 9): the IGF-I receptor (typeIIGF receptor), the IGF-I1 receptor (type I1 IGF receptor), and the crine growth and motility factor in human rhabdomyosarcoma insulin receptor. The insulin and the type I IGF receptors cell lines (5). The mitogenic activity was inhibited by a-IR3, a blocking monoclonal antibody (Ab) specific for the type I bindtheir respective ligands with high affinity andboth heterologous peptides with lower affinity. These receptors are IGF receptor. On the other hand,the specific IGF-11-mediated increase in rhabdomyosarcoma cell motility was not affected * The costs of publication of this article were defrayed in part by by preincubation with a-IR3. These findings prompted us to the payment of page charges. This article must therefore be hereby further analyze the mechanism of transduction of the IGF-IImarked “advertisement” in accordance with 18 U.S.C. Section 1734 induced motility signal in rhabdomyosarcoma cells. azp2
solely to indicate this fact. 8 To whom correspondence should be addressed:Pediatric Branch, EXPERIMENTALPROCEDURES National Cancer Institute, Bldg. 10, Rm. 13N240, Bethesda, MD Materials-Human recombinant IGF-I and IGF-I1 werepurchased 20892. Tel.: 301-496-1543; Fax: 301-402-0575. $$ Supported by Dr. Mildred Scheel Stipendium (Deutsche Kreb- from Bachem, Torrance, CA. Human recombinant insulin was purshilfe e.V.). chased from Collaborative Research, Bedford, MA.[Leu*’]IGF-II The abbreviations used are: IGF, insulin-like growth factor; CI- synthesis and specificities have been described previously (24, 25). MPR, cation-independent mannose 6-phosphate receptor; Ab, anti- Goat nonimmune serum was purchased from Vector Laboratories body; DMEM, Dulbecco’s modified Eagle’s medium; BSA, bovine Inc., Burlingame, CA. The goat anti-type I1 receptor Ab has been serum albumin; MTT, 3-[4,5-dimethylthiazol-2-yl]2,5-diphenyltetra-previously described (26). The purified IgG to the type I1 receptor zolium bromide; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesul- was obtained with passage of the whole antiserum through a G column fonic acid. (Pierce, Rockford, IL). To functionally verify that the IgG fraction
9000
The Type
11IGF Receptor Mediates Motility
in Rhabdomyosarcoma
9001
isolated contained an intact and active antibody against the human human serum albumin, 10 mM mannose 6-phosphate (1.0 ml/plate) IGF-II/CI-MPR we assessed its ability to inhibit theendocytosis of for 30 minon ice. The cells were then washed three times with &glucuronidase compared with thewhole antiserum. We found that phosphate-buffered saline(137 mM NaC1,3 mM KCl, 8 mM NazHP04, 1.4 mM KH2P04, pH7.2) and 0.05% saponin. They were incubated the optimal inhibition (90%) occurreda t a concentration of purified IgG of 150 pg/ml (data not shown). The monoclonal Ab a-IR3 was with 20,000 units/ml recombinant human @-glucuronidase in 50 mM provided by Drs. Steven Jacobs and Fred C. Kull (Burroughs Well- Hepes, pH 7.5, 150 mM NaCl, 5 mM P-glycerophosphate, 0.5% human come, Research Triangle Park, NC).Iz5I-IGF-II was purchased from serum albumin, 0.5% saponin with or without 10 mM mannose 6phosphate for 3 h on ice. The cells were then washed five times with Amersham Corp. All other chemicals were reagent grade. Cell Culture-Human rhabdomyosarcoma cells (RD, Ref. 27) were phosphate-buffered saline and0.05% saponin andwere solubilized in 1% sodium deoxycholate. The solubilized cells were assayed for @grown in DMEM supplemented with 10% fetal bovine serum and glucuronidase (33) and protein. Under these conditions (pH 7.5), the antibiotics in a 37 "C humidified incubator with 6% CO,. For the cation-dependent 46-kDareceptordoes not contribute to enzyme motility assay the cells were harvested when 80% confluent by a brief exposure to 0.25% trypsin, followed by a 2-h incubation period a t binding. Affinity Cross-linking Studies andRadioreceptor Assays-RD memroom temperature in order allow to for the regenerationof the surface receptors, and thenwere resuspended in 0.1% bovine serum albumin branes were prepared from approximately 4 X lo7 RD cells, by (BSA)/DMEM. For the cell growth experiments RDcells were trans- resuspending the pellets in10 volumes of buffer 1 (10 mM Tris-HC1, ferred in the serum-freemedium N2E (4) andpassaged a t least once 0.25 M sucrose, 1.5 mM EDTA, 0.5 mM dithiothreitol, 1 mM benzin N2E prior to study. Under these conditions RD cells grow auton- amidine, and 0.002% soybean trypsin inhibitor) and homogenized with a glass Teflon homogenizer (Thomas Scientific, Swedsboro,NJ). omously via an IGF-I1 autocrineloop (5). Cell Motility Assay-The assay used to determinecell motility has The homogenate wascentrifuged at 1,000 X g for 10 min.The supernatant was adjusted to 0.1 M NaCl and spun at 160,000 X g for been described in detail previously (28). Briefly,nucleoporemembranes (8-pm pore size) were coated with gelatin (100 pg/ml in 0.16 30 min. The pellets were resuspended in50 mM Tris, and the protein M acetic acid) to promote cell attachment. Human recombinant IGF- concentration was determined using the Bradford method (30). Aliquots of the membranes were stored at -70 "C until use. For the I1 and [Leuz7]IGF-II were diluted to the stated concentrations0.1% in BSA/DMEM and used as chemoattractants in the Boyden chamber cross-linking assay, 50 pg of RD membranes were preincubated with assays. The motility assay was incubated for 4 h, fixed, and stained. cold insulin, IGF-I, IGF-11, goat Ab, or goat nonimmune serum at the indicated concentrationsfor 30 min at 23 "C. lZ5I-IGF-II or '251-insulin The 0.1% BSA/DMEM served as the negative controltoassess unstimulated motility. All experiments were performed in triplicates (300,000 cpm) was added and the samples incubated overnight at and were repeated a minimum of three times. Motility was quantified 0 "C. Affinity cross-linking and sodium dodecyl sulfate-polyacrylby counting the number of cells per 5 high power fields (500x1 by amide gel electrophoresis were performed by the methodof Massague two independent investigators (C. P. Minniti and E. C. Kohn). The and Czech (31), using a prepoured 4-20% Tris/glycine gradient mincells were preincubated for 1 or 2 h to evaluate the ability of a-IR3 igel (Novex, Encinitas, CA), and the discontinuousbuffer system of or the anti-type I1 receptor Ab to inhibit RD cell motility. a-IR3 (1 Laemmli (32). pg/ml, the concentration that maximally inhibited cell growth in Fortheradioreceptorassay, 50 pg of membraneprotein were serum free medium) (5) was used in the cell motility experiments. incubated with 10 pM iodinated peptide, in the presence or absence Since the anti-typeI1 Ab did not inhibit RDcell growth (see below), of unlabeled peptide overnight (16 h) a t 4 "C. The pH, time, and the dilution of Ab that maximally inhibited the uptake of lysosomal temperature conditionswere determined to be optimal in preliminary enzymes of human fibroblasts (1:lOO) was used. A dose curve of the assays. The membranes were washed in several volumesof cold buffer, inhibition of IGF-11-induced motility by the anti-type I1 Ab, ranging and bound radioactivity was determined in an automatic y counter. from 1:50 to 1:1000, also showed that 1:lOO was the optimal dilution Nonspecific counts were determined in the presence of excess unla(data not shown). At this dilution of antibody no binding of IGF-I1 beled peptide: 200 ng/ml IGF-I. to the typeI1 receptor could be detectedby cross-linking experiments Statistical Analysis-Data were analyzed with a two-tailed Stu(see "Results" and Fig. 5). dent's t test, using the software program Statview (Abacus Concept, In experiments testing the purified IgG fraction, RD cells were Inc.). preincubated with 150 pg/ml of either anti-type I1 receptor IgG or nonimmune goat IgG and then assayed for a motility response to RESULTS [Leu"]IGF-11. A dose of 10"' M [LeuZ7]IGF-IIwas used in these sets [ L ~ u ~ ~ I I G F Stimulates I I the Motility of Rhabdoof experiments in orderto avoid any possibility of binding to the type myosarcoma Cells-In a previous report we have shown that I IGF-receptors, asassessed by the binding data(see below). Mitogenesis Assays-Exponentially growing RD cells were plated IGF-I1 stimulates RD cell motility and this activity was not into 96-well dishes (5 X lo3 cells/well) in triplicate in N2E medium inhibited by blocking the type I receptor (5). Since IGF-I1 (4).After cells adhered for 12-18 h, the anti-type I1 receptor Ab or binds to the type I and the type I1 IGF receptor, to further equivalentamounts of goat nonimmuneserum was added to the explore the mechanism by which IGF-I1 stimulates the motilappropriate wells. The anti-typeI1 receptor Ab and the control serum were re-added after 72 h. Cell number was assessed with the MTT ity of rhabdomyosarcoma cells, we used [LeuZ7]IGF-IIas a chemoattractant in a modified Boyden chamber assay. This (3-[4,5-dimethylthiazol-2-yl]2,5-diphenyltetrazolium bromide) assay (29). This assay measures the reduction by living cellsof a tetrazolium is an analog of IGF-I1 that binds selectively the type I1 salt, MTT (Sigma), to a blue-colored formazan product (4). Briefly, receptor with an affinity 100-fold greater than the type I 5 X 10'' cells were plated in 96-well culture dishes in 200 $1 of the receptor (24). [LeuZ7]IGF-IIwas able to stimulate RD cell indicated medium. At the indicated times the medium was removed and 100 p1 of M T T were added (0.5 mg/ml final concentration) and motility at a concentration as low as 10"' M (Fig. 1).The the incubation continuedfor 3 additional h at 37 "C. The precipitated difference in cell number between unstimulated and lo-'' M formazan was solubilized with an equal volume of propanol and the [Leuz7]IGF-11-stimulatedRD cells estimated in eight individabsorbance determined at 570 and 690 nm using an enzyme-linked ual experiments was statistically significant ( p = 0.0001). At immunosorbent assay reader (Titertek Multiskan MCC/340). After this concentration [LeuZ7]IGF-IIdoes not bind to the type I correction for the nonspecific absorbance a t 690 nm, the absorbance receptor as shown by the radioreceptor binding assay (Fig. 2). at 570 nm was found to be linearly proportional to cell number over Cold IGF-I displaces the lZ5I-IGF-Itracer better than IGF-11, the range of cell concentrations examined in our experiments (data not shown). The cell number in each well was determined from a and [Leuz7]IGF-I1 does not displace labeled IGF-I at concenstandard curve relatingabsorbanceto cell number,generated by trations as high as 20 nM. Each point representsthe mean of plating a known number of cells intriplicate wells. Treated and duplicate data points. This finding, together with our earlier untreated cells were examined microscopically prior to the addition observation that theIGF-11-mediated motility was not inhibof the dye, to assess cell viability. Each data point represents the itable by preincubation with a-IR3, suggests that rhabdomean k S.D. of triplicate determinations. myosarcoma cell motility response may be mediated, at least /3-Glucuronidme Binding to RD Cells-RD cells and human fibroblasts were grown as monolayers in 35-mm cell culture dishes. For in part, through the type I1 receptor. Antibodies to the Type 11 IGF Receptor but Not the TypeI binding assays, cells were permeabilized with 0.25% saponin in 50 mM Hepes, pH 7.0, 150 mM NaCl, 5 mM &glycerophosphate. 0.5% Receptor Inhibit [LeuZ7]IGF-II- and IGF-11-induced Motil-
9002
The Type11IGF Receptor Mediates Motilityin Rhabdomyosarcoma 300I -
A. 300
W
untreated
H
MOPC
,-
?
f
200
c
-e 0
s
-I-
.01
0
-rl
1
100
10
1
100
[leu-27 IGF-Ill nM
FIG. 1. Rhabdomyosarcoma cells migrate to [Leu2']IGF-II. RD cells were harvested with 0.25% trypsin and allowed to regenerate the surface receptors for 2 h at room temperature, and 1.1 X lo5 cells were placed in the upper wells of the modified Boyden chamber. The motility assay was performed as described under "Experimental Procedures." Increasing concentrations of [Leu2']IGF-II ranging from 0.01 to 100 nM, or plain 0.1% BSA/DMEM, used as control or unstimulated motility, were placed in the lower wellsas chemoattractant. After 4 h of incubation at 37 "C, the chamber was opened and the number of cells that had migrated in response to the ligand counted. Data are presented as percent of control, obtained by dividing the mean of triplicate determinations of individual data points to the mean of the triplicate determination of the controls. Each data point is a mean f S.E. of two individual experiments.
B.
W
anti-type-ll Ab
unstimulated
"
I
0.1
10
8
'
.
.. 100
Peptide concentration (nM) FIG. 2. Radioreceptor assay on RD membranes. Rhabdomyosarcoma cell membranes (50 pg of protein) were incubated for 18 h at 4 "C with lz5I-IGF-I in the presence or absence of varying concentrations of IGF-I (A),IGF-I1 (O),or [Leu2']IGF-II (+). Results shown are means of duplicate determinations and are representative of two independent experiments. B and Bo are specific binding in the presence and absence, respectively, of excess unlabeled ligand.
Untreated Goal serum
10-BM
IGF-II
FIG. 3. Motility response of rhabdomyosarcoma cells pretreated with anti-type I or anti-type I1 receptor Ab. RD cells were harvested, rested, and then incubated in serum freemedium with the appropriate Ab. The incubation with a-IR3, shown in panel A , did not inhibit IGF-11- nor [Leu2']IGF-II-induced motility. The data presented in panel A refer to an individual representative experiment. The incubation with the anti-type I1 receptor antiserum, shown in panel B, totally abolished the motility response of RD cells to lo-' M IGF-11. The difference between treated and untreated cells was statistically significant ( p = 0.0095 on a two-tailed Student's t test), while the slight decrease observed with preincubation with goat serum was not significant ( p = 0.35). The data shown in panel B are the mean k S.E. of three individual experiments.
binding proteins in theassay, we preincubated RD cells with the purified IgG antibody fraction to the type I1 receptor or with pure goat nonimmune IgG. We thenassayed for a motility response to 10"' M [LeuZ7]IGF-II. The results of a representative experiment are reported in Fig. 4 and show that the purified Ab was able to inhibitcompletely the [LeuZ7]IGF-XIity-To further determine the specific receptor involved in induced stimulation of RD cell motility, while it had noeffect IGF-XI-induced motility response, we analyzed the ability of on unstimulated motility. The nonimmune goat IgG had no antibodies to the typeI and type I1 receptor to alter IGF-II- effect. The Antibody to the Type 11Receptor Does Not Inhibit RD mediated motility. As shown in Fig. 3 (panel B ) , the Ab to the type I1 receptor completely inhibited IGF-11-induced mo- Cell Growth-The IGF-I1 autocrinegrowth loop is specifically tility, while the control goat serum had no effect ( p = 0.0095 mediated through the IGF-Ireceptor ( 5 ) .To demonstrate that and 0.35, respectively). On the other hand, as shown in Fig. the IGF-I1 autocrine motilityloop is different and independ3A, 1 pg/ml a-IR3, a blocking Ab to the type I receptor, did ent of its autocrine growth loop, we assayed the ability of the not inhibit the motility response elicited by either IGF-I1 or anti-type I1 receptor Ab toinhibitRD cell growth. The I1 receptor was added to RD cells grown [LeuZ7]IGF-II.MOPC-21, a matched mouse IgG (1 pg/ml), antiserum to the type was used as the control.To further address thespecificity of in serum free medium to a final concentration of 1:100, after the response observed and exclude the interference of IGF 12 and 72 h of incubation. Cell number was assessed using
T h e Type 11 IGF Receptor Mediates Motility in Rhabdomyosarcoma T
Unstirnulated
[leuZ7]lGF-il10-10M
FIG. 4. The purified antibodyto t h e t y p eI1 receptor inhibits [Leu"]ICF-I1 motility. RD cells were harvested, rested, and then nonimmunegoat IgG incubated a t 37 "C for 2hwith150pg/ml (stippled bar), with plain 0.1% RSA/DMEM (solid bar), or with 150 pg/ml purified IgG anti-type I1 receptor (hatched bar).lo-"' M [Leu"] IGF-I1 was used as a chemoattractant. Purified IgG to the type I1 receptor (hatchedbar) totally abolished RD motilityresponse to [Leu"]IGF-I1 but had no effect on the unstimulated motility (solid bar). Nonimmune goat IgG (stippled bar) hadno effect on either motility response. Each bar represents the mean & S.D. of triplicate determinations.
TABLE I P-Glucuronidase binding to permeabilized RD cells or human fibroblasts Cells were permeabilized as described under "Experimental Procedures," and then the binding of @glucuronidase was assayed. RD cell /3-glucuronidase binding to the mannose-6-phosphate receptor was compared with the binding obtained when using human fibroblasts.
9003
as much P-glucuronidase as human fibroblasts at pH 7.5, at which pH all of the enzyme binding is due to theIGF-II/CIMPR. The ability of labeled IGF-I1to bind the type I1 receptor and the ability of type I1 receptor antibody to block this interaction were examined by cross-linking experimentsusing labeled IGF-11. The Ab used in the inhibitionof motility was able tocompete for IGF-I1binding on the type I1 receptor. As demonstrated in Fig. 5, '"I-IGF-I1 strongly labels both the a subunit of the type I IGF receptor at 130 kDa and the typeI1 IGF receptora t 270 kDa (lane 9 ) . Excess cold IGF-I1 (lane 8 ) totally displaces the tracer from both bands. The anti-type I1 receptor Ab a t dilutions of 1:200, 1:100, and 1:50 specifically attenuated the 270-kDa type I1 receptor band (lunes 3-5, respectively). Of note, the concentration of anti-serum used to inhibit IGF-11- and [Leuz7]IGF-11-induced motility (1:lOO) was able to completely attenuate IGF-I1 binding to the type I1 receptor as assayed by cross-linking studies. Insulin and IGF-I did not displace the IGF-I1 tracer from the 270-kDa band, while they both attenuated the 130Kd band corresponding to the type I receptor (lanes 6 and 7). Moreover, goat nonimmune serum at 1:lOO and 1:50 dilution did not displace the IGF-I1 from the type I1 receptor 270-kDa band (lunes I and 2 ) . Similar resultswere obtained when the purified IgG fraction of the antibody was used (data not shown). DISCUSSION
Rhabdomyosarcomas areaggressive embryonal tumorswith a high propensity to metastasize early (34). The ability to transit out of the capillary wall and into the stroma is an essentialrequirement for themetastatic process (35). We have previously documented the role of IGF-I1 in the growth and locomotion of rhabdomyosarcoma cells (5), and we have Mannose 6-phosphateHuman fibroblasts demonstrated that the mitogenic signal was mediated specifCell line specific binding binding value ically by the type I IGF receptor in an autocrinegrowth loop. unitsjmg protein ?& In this study we sought to ascertain which of the IGFreceptors Human fibroblasts 1467 100 mediated IGF-11-induced RD cell locomotion. RD cells 865 59 The availability of antibodies thatselectively recognize and block IGF receptor subtypes, and more recently the in vitro the MTT assay after 24, 48, 72, 96, and 120 h of incubation. synthesis of IGF-I1 analogs, like [LeuZ7]IGF-II, withdifferIn the controlwells nonimmune goat serumwas added at the ential affinity for one of the two major IGF receptors has same time toa final dilution of 1:lOO. We did notobserve any greatly enhanced our ability to attributespecific functions to inhibition of RD cell growth but a slightincreasein cell each individual receptor. Using these toolswe have presented number(datanotshown). A possible explanation for the data that strongly implicate that IGF-I1 stimulates motility through the IGF-II/CI-MPR creating a potential autocrine observed enhancement of RD cell growth could be the increased availability of the IGF-I1 peptide for binding to the motility loop. There are several lines of evidence to support type I receptor, due to the functional inactivationof the type our conclusions. First, preincubationwith a-IR3 was not able I1 receptor. This hypothesis is further corroborated by the kD results of the cross-linking experiments (see below) which show an increase in intensity of the band corresponding to the a subunit of the type I receptor when the membranes -200 were incubated in the presence of the Ab anti-type I1 receptor. We repeated the same experiment using the purified IgG fraction of the antiserum with a consistent lack of effect on e RD cell growth (data not shown). These findings are in sharp contrast to those obtained when -97.5 n-IR3 was used under the same experimental conditions(5), 1 2 3 4 5 6 7 8 9 which supports the hypothesis that the two IGF-I1 effectsare FIG. 5. Cross-linking of t h e t y p e I1 receptor. RD cell memmediated by different receptors. branes were incubated under reducing conditions with 300,000 cpm RD Cells Express the Type II Receptor and the Anti-type II of "'I-IGF-I1 (lane 9 ) and the ligands indicated in the figure: goat nonimmune serum a t 1:lOO and 1:50 dilution (lanes I and Z),antiReceptor Ab Inhibits the Binding of IGF-II to This ReceptorIn most cells the type I1 receptor is predominantly located in type I1 receptor Ab a t 1:200, 1:100, and 1:50 dilution (lanes 3-5), 1 cold insulin (lane 6 ) , 200 ng/ml cold IGF-I (lane 7), and 200 the intracellular compartment(33). T o demonstrate that RD pg/ml ng/ml cold IGF-I1 (lane 8). Lane 9 indicates the labeling of both cells possess a type I1 receptor, we examined enzyme binding receptors by ""I-IGF-11. The upper solid arrow indicates the type I1 and IGF-I1 binding. Table I demonstrates the binding of P- IGF receptor a t 270 kDa, the lower clear arrow indicates the subunit glucuronidase to permeabilized RD cells. RD cells bind 50% of the type I IGF receptor a t 130 kDa. (Y
9004
The Type
11IGF Receptor Mediates Motility
in Rhabdomyosarcoma
t o block the IGF-11- or [Le~~~]IGF-II-induced motility. Sec- messenger pathway for the IGF-II/CI-MPR (18, 19)using this unique biological assay. ond, an antibody that functionally inactivates the human IGF-II/CI-MPR specifically inhibited theIGF-11- and [Leuz7] Acknowledgments-We thank Dr. Daisy De Leon and Dr. Peter IGF-11-induced motility, while it had no effect on insulinNissley for helpful discussions and encouragement. induced motility (data not shown) or unstimulated motility. REFERENCES T h i r d , [ L ~ u ~ ~ I I G was F - I Iable to elicit a motility response a t 1. Rinderknecht, E., and Humbel, R. E. (1976) Proc. Natl. Acad. Sci. U. S. A. doses that bind only the type I1 receptor. Since insulin is 73,2365-2369 known to stimulate motility in some tumor cells (36, 37), 2. Nissley S. P., and Rechler M. M. (1986) Horm. Protein Pe t 12,127 203 3. Rinderinecht, E., and Humbel, R. E. (1978) FEES Lett. 8b),'283-289including rhabdomyosarcoma (data not shown), and IGF-I1 4. El-Badry, 0. M., Romanus, J. A,, Helman, L. J. Cooper M. J., Rechler, M. M., and Israel, M. A. (1989) J. Clin. Inuest. 5 4 , 8291839 can bind to the insulin receptor, thepossibility that IGF-II5. El Badry, 0.M., Minniti, C. P., Kohn E. C., Houghton, P. J., Daughaday, induced motility is mediated by this receptor mustbe considW. H., and Helman, L. J. (1990) Cell Growth & Differ. 1,325-331 ered. However, the doses of [LeuZ7]IGF-II and IGF-I1 that are6. Yee, D., Cullen, K. J., Paik, S., Perdue, J. F., Hampton, B., Swartz, A,, Lippman, M. E., and Rosen, N. (1988) Cancer Res. 48,6691-6696 effective in stimulating the motility of RD are in the nano7. Daughaday, W. H. (1990) Endocrinology 1 2 7 , l - 4 8. Czech, (1989) Cell 5 9 , 235-238 molar range, levels that arehighly unlikely to bind the insulin 9. Rechler,M.M.P.M., and Nissley, S. P. (1985) Annu. Reu. Physiol. 4 7 , 425-442 receptor. Inaddition,cross-linkingexperiments using the 10. Ellis, L., Clauser, E., Morgan, D. O., Edery, M., Roth, R. A., and Rutter, W. J. (1986) Cell 4 5 , 721-732 antibody to the type I1 receptor that blocks IGF-11-induced 11. Rosenfeld, R. G., Conover, C. A,, Hodges, D., Lee, P. D. K.,Misra, P., Hintz, R. L., and Li, C. H. (1987) Biochem. Biophys. Res. Commun. 1 4 3 , motility demonstrate the inability of this antibody to displace 199-205 insulin from the insulin receptor (data notshown). 12. Ewton, D. Z., Falen, S. L., and Florini, J. R. (1987) Endocrinology 1 2 0 , 115-123 It has recently been shown that virtually all IGFs inblood, 13. MacDonald, R. G., Pfeffer, S. R., Coussens, L. Te per, M. A., Brocklebank, other extracellular fluids, and cell culture media are noncoC. M., Mole,.J. E., Anderson, J. K., Chen, E., Ezech, M. P., and Ullrich, A. (1988) ScLence 2 3 9 , 1134-1137 valently bound to several distinct forms of specific carrier 14. Oshima, A,, Nolan, C. M., Kyles, J. W., Grubb, J. H., and Sly, W. S. (1988) J. Biol. Chem. 2 6 3 , 2553-2562 proteins (38). Although the role of the IGF binding proteins, P., Dahms, N. M., Breitmeyer, J., Chirgwin, J. M., and Kornfeld, S. aside fromgreatly extending the plasma half-life of these 15. Lobel, (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 2233-2237; Correction (1987) Proc. Acad. Sci. U. S. A. 8 4 , 7523 hormones, is not fully understood, they have the capacity to 16. Lobel, P.,Natl. Dahms, N.M., and Kornfeld, S. (1988) J. Biol. Chem. 263,2563modulate IGFactions. In mostcases, the IGF-binding protein 2570 W., Blickenstaff, G. D., Sklar, M. M., Thomas, C. L., Nissley, S. P., complex doesnot bind tocell receptors and is inactive. There- 17. Kiess, and Sahagian, G. C. (1988) J. Biol. Chem. 263,9339-9344 18. Nishimoto, I. Hata, Y., Ogata, E., and Kojima, I. (1987) J. Biol. Chem. fore one could speculate that the inhibition of the IGF-II262,12120-12126 induced motility couldbe secondary to the presence of binding 19. Okamoto, T., Toshiaki, K., Murayama, Y., Ui, M., Ogata, E., and Nishimoto, I. (1990) Cell 62, 709-717 proteins in thegoat antiserum. This isunlikely to be the case, 20. von Fi ra, K., and Hasilik, A. (1986) Annu. Reu. Biochem. 5 5 , 167-193 since 1) the purified IgG fraction of the antiserum has the 21. Hari, TPierce, s. B., Morgan, D. O., Sara, V., Smith, M. C., and Roth, R. (1987) EMBO J. 6,3367-3371 same inhibitoryeffect as the whole antiserum and2) the goat 22. Tally, M., Li, C. H., and Hall, K. (1987) Biochem. Biophys. Res. Commun. 148,2,811-816 nonimmune serum has no effect on RD cell motility. J.. Gavin. J. R.. and Hammermann. M. R. (1986) . . J. Biol. Chem. To date onlybombesin (39) and the IGFs (35, 36) have 23. Mellas. 261; 14437-14442 24. Beukers. M. W.. Oh. Y.. Zhane. H.. Line. N.. and Rosenfeld. R. G. (1991) been reported tohave the ability to stimulate both growth the Endocrimlo y 1 2 8 , 1201-1503 and the motility of tumor cells. In both cases, the mitogenic 25. Ling,N.Escg F. BohlenP.,Brazeau,P.,Wherenberg, W. B.,and Guillemin R.'(1484) Proc.'Natl. Acad. Sci. U. S. A . 81,4302-4306 and themotility functions appear to be mediated by the same 26. Nolan, C. hi., Creek, K. E., Grubb, J. H., and Sly, W. S. (1987) J. Cell. Biochem. 3 5 , 137-151 single receptor, the bombesin receptor (40) and the type I IGF 27. McAllister, R. M., Melnyk, J., Finkelstein, J. Z., Adams, E. C., Jr., and receptor (36). We believe that this report documents that a Gardner, M. B. (1987) Cancer 24,520-526 M. L., Guiris, R., Liotta, L. A,, andShiffmann,E. (1987) single, autocrine produced peptide can mediate two different 28. Stracke, BLochem. Bio hys E s . Commun. 146,339-345 functions through two distinct and independent receptors in 29. Denizot, F., ancflang, R. (1986) J. Immunol. Methods 89,271-277 30. Bradford M. M. (1976) Anol. Biochem. 72,248-254 the same malignant cell type. Growth and motility of rhab- 31. Massam;, J., and Czech, M. P. (1982) J. Biol. Chem. 257,5038-5045 domyosarcoma cells appear to be independentof each other, 32. Laemmli, U. K. (1970) Nature 227,680-685 33. Fischer, H. D., Gonzalez-Noriega, A,, and Sly, W. S. (1980) J. Biol. Chem. being mediated by the typeI and typeI1 receptor, respectively. 255,5069-5074 B. Jr., Hays, D. M., Tefft, M., and Triche,T. J. (1981) in Principles This provides new insight into the independence of tumori- 34. Raney, and Practice of Pediatric Oncolog (Pizzo, P. A,, and Poplack, D. G., eds) pp. 635-658, Lip incott Co., Phlgdelphia genesis and metastatic progression, and our results provide Liotta, L. A. (1986PCancerRes. 4 6 , 1-7 strong evidence for a new model of diversification of responses 35. 36. Stracke, M. L., Kohn, E. C., Aznnavoorian, S. A,, Wilson, L. L., Salomon, D., Krutzsch, H. C., Llotta, L. A., and Shiffmann, E. (1988) BLochem. and maximal utilization of the same peptide by a neoplastic B i o p p Res. Commun. 163,1076-1083 cell. 37. Kohn, . C., Francs, E. A,, Liotta, L. A,, and Shiffmann, E. (1990) Int. J. Cancer 46 287-292 Moreover, the characterizationof a functional assayfor the 38. Rechler, M. M., andNissle S P (1990) in Handbook o Ex erimental Pharmacology (Sporn, M. andRoberts, A. B., eds) Vot95{, pp. 317IGF-II/mannose 6-phosphate receptor may allow investiga343, S rin er-Verlag, Berlln tors to analyze the ability of IGF ligands to mediate their 2 Sliffmann E., Terranova, V., andPert, C. B. (1985) Clin. Ruff, 39. fmmu&l. fmmunoiathol. 37,387-396 effects through this receptor.Finally, we are currently at40. Cuttitta, F., Carne D. N Mulshine J., Mood T. W., Fedorko, J., Fischler, A,, and hinna, J : D. (1985) hature 31$823-826 tempting to test the suggestedG protein-mediated second ~
~~
~
'
'
B.,
"
'
